<%BANNER%>

Cross-Layer Design of Networking Protocols in Wireless Local Area Networks and Mobile Ad Hoc Networks


PAGE 4

Firstandforemost,Iwouldliketoexpressmysinceregratitudetomyadvisor,Pro-fessorYuguangFang,forhisinvaluableadvice,encouragementandmotivationduringthecourseofthiswork.Thisdissertationwouldnothavebeenpossiblewithouthisguid-anceandsupport.Ialsothankhimforhisphilosophicaladviceonbothmyacademicandnonacademiclife,whichmadememoremature,scholasticallyandpersonally.IthankProfessorsShigangChen,JoseFortes,PramodKhargonekarandSartajSahniforservingonmysupervisorycommitteeandfortheirvaluablesuggestionsandconstruc-tivecriticism.ThanksalsogotoProf.JohnShea,Prof.TanWongandProf.DapengWu,fortheirmanyconstructivesuggestionsandadvice.ManythanksareduetomycolleaguesDr.XiangChenandJianfengWangfortheircollaboration.IalsothankDr.YounggooKwon,Dr.WenjingLou,Dr.WenchaoMa,Dr.WeiLiu,Dr.Byung-SeoKim,Dr.XuejunTian,Dr.SungwonKim,Dr.JaeSungLim,YuZheng,YanchaoZhang,ShushanWen,XiaoxiaHuang,YunZhou,JingZhao,ChiZhang,FrankGoergen,PanLi,FengChen,ShanZhang,RongshengHuangandmanyothersatUniversityofFloridafortheyearsoffriendshipandmanyhelpfuldiscussions.Lastbutnotleast,Ioweaspecialdebtofgratitudetomyparentsandmybrothers.Withouttheirselessloveandsupport,IwouldneverimaginewhatIhaveachieved. iv

PAGE 5

page ACKNOWLEDGMENTS ................................ iv LISTOFTABLES ................................... xii LISTOFFIGURES ................................... xiii ABSTRACT ....................................... xviii CHAPTER 1INTRODUCTION ................................ 1 1.1Motivation ................................. 1 1.2OrganizationoftheDissertation ..................... 2 2PERFORMANCEOFTHEIEEE802.11DCFPROTOCOLINWIRELESSLANS ...................................... 8 2.1Introduction ................................ 8 2.2Preliminaries ............................... 10 2.2.1DistributedCoordinationFunction(DCF) ............ 10 2.2.2SystemModeling ......................... 11 2.3TheProbabilityDistributionoftheMACLayerServiceTime ...... 12 2.3.1MACLayerServiceTime ..................... 12 2.3.2ProbabilityGeneratingFunctions(PGF)ofMACLayerServiceTime ............................... 13 2.3.3TheProcessesofCollisionandSuccessfulTransmission ..... 15 2.3.4DecrementProcessofBackoffTimer ............... 16 2.3.5MarkovChainModelfortheExponentialBackoffProcedure .. 17 2.3.6GeneralizedStateTransitionDiagram .............. 18 2.3.7ProbabilityDistributionModeling ................ 20 2.3.8DerivationofTransmissionProbability .............. 23 2.4QueueingModelingandAnalysis ..................... 25 2.4.1Problemformulation ........................ 25 2.4.2Thesteady-stateprobabilityoftheM/G/1/Kqueue ........ 26 2.4.3ConditionalCollisionProbabilitypcandDistributionofMACLayerServiceTime ....................... 27 2.4.4PerformanceMetricsoftheQueueingSystem .......... 27 2.4.5Throughput ............................ 27 2.4.6NumericalResults ......................... 28 v

PAGE 6

......................... 29 2.5.1SimulationEnvironments ..................... 29 2.5.2ProbabilityDistributionofMACLayerServiceTime ...... 30 2.5.3ComparisonofM/G/1/KandM/M/1/KApproximationswithSimulationResults ....................... 30 2.6Conclusions ................................ 33 3HOWWELLCANTHEIEEE802.11DCFPROTOCOLSUPPORTQOSINWIRELESSLANS ............................... 35 3.1Introduction ................................ 35 3.2Preliminaries ............................... 37 3.2.1OperationsoftheIEEE802.11 .................. 37 3.2.2RelatedWork ........................... 38 3.3AnalyticalStudyoftheIEEE802.11 ................... 40 3.3.1MaximumThroughputandAvailableBandwidth ......... 40 3.3.2DelayandDelayVariation .................... 47 3.3.3PacketLossRate .......................... 54 3.4SimulationStudyoftheIEEE802.11 ................... 56 3.4.1SimulationConguration ..................... 56 3.4.2SimulationResults ......................... 58 3.5Discussions ................................ 60 3.5.1ImpactofFadingChannel ..................... 60 3.5.2ImpactofPrioritizedMAC .................... 61 3.6Conclusion ................................. 61 4ACALLADMISSIONANDRATECONTROLSCHEMEFORMULTIME-DIASUPPORTOVERIEEE802.11WIRELESSLANS ........... 62 4.1Introduction ................................ 62 4.2Background ................................ 65 4.2.1OperationsoftheIEEE802.11DCFProtocol .......... 65 4.2.2QoSRequirementsforMultimediaServices ........... 66 4.3ChannelBusynessRatio .......................... 67 4.3.1DenitionofChannelBusynessRatio .............. 67 4.3.2Channelbusynessratio:anaccuratesignofthenetworkutilization 68 4.3.3MeasurementofChannelBusynessRatio ............. 71 4.4CARC:CallAdmissionandRateControl ................ 71 4.4.1DesignRationale ......................... 72 4.4.2CallAdmissionControl ...................... 74 4.4.3RateControl ............................ 76 4.5PerformanceEvaluationofCARC .................... 79 4.5.1SimulationConguration ..................... 79 4.5.2SimulationResults ......................... 80 4.6Discussions ................................ 85 4.6.1ImpactofFadingChannel ..................... 85 vi

PAGE 7

.................... 86 4.7Conclusion ................................. 87 5DISTRIBUTEDFAIRANDEFFICIENTRESOURCEALLOCATIONWITHQOSSUPPORTOVERIEEE802.11WLANS ................ 88 5.1Introduction ................................ 88 5.2DesignRationale ............................. 92 5.2.1EfciencyandQoS ........................ 92 5.2.2Fairness .............................. 94 5.3DistributedResourceAllocation(DRA) ................. 95 5.3.1BasicFramework ......................... 96 5.3.2FairnessSupport .......................... 100 5.3.3QoSSupport ............................ 100 5.3.4MultipleChannelRatesSupport ................. 102 5.4ConvergenceAnalysis ........................... 102 5.4.1ConvergenceofMultiplicative-IncreasePhase .......... 102 5.4.2ConvergencetoFairnessEquilibrium ............... 105 5.4.3Discussion ............................. 109 5.4.4ParameterSelection ........................ 110 5.5PerformanceEvaluation .......................... 110 5.5.1SimulationSetup ......................... 110 5.5.2ChannelBusynessRatioThreshold ................ 111 5.5.3Fairness .............................. 112 5.5.4Efciency,DelayandCollision .................. 115 5.5.5QualityofService ......................... 116 5.6RelatedWorkandDiscussions ...................... 119 5.7Conclusions ................................ 121 6PHYSICALCARRIERSENSINGANDSPATIALREUSEINMULTIRATEANDMULTIHOPWIRELESSADHOCNETWORKS ........... 123 6.1Introduction ................................ 123 6.2OptimumCarrierSensingRange ..................... 127 6.2.1AggregateThroughputandSINRattheWorstCase ....... 127 6.2.2MaximumThroughputandOptimumCarrierSensingRangeun-derShannonCapacity ...................... 130 6.2.3MaximumThroughputandOptimumCarrierSensingRangeun-dertheDiscreteChannelRatesoftheIEEE802.11 ...... 131 6.2.4ImpactofRandomTopology ................... 133 6.2.5TradeoffbetweenExposedTerminalProblemandtheHiddenTerminalsProblem ....................... 134 6.2.6CarrierSensingRangeandStrategiesforBidirectionalHand-shakes .............................. 136 6.2.7OptimumCarrierSensingRange ................. 140 vii

PAGE 8

................................ 140 6.3.1HowtoSettheCarrierSensingThresholdforMultirate802.11MACprotocol .......................... 140 6.3.2HowtoChooseNextHops,ChannelRatesandSettheCarrierSensingThresholdforMultihopFlows ............. 142 6.4SimulationStudies ............................ 149 6.4.1NS2ExtensionsandSimulationSetup .............. 150 6.4.2OptimumCarrierSensingRange ................. 150 6.4.3SpatialReuseandEnd-to-EndPerformanceofMultihopFlows 153 6.5Conclusions ................................ 154 7ADUAL-CHANNELMACPROTOCOLFORMOBILEADHOCNETWORKS 156 7.1Introduction ................................ 156 7.2Background ................................ 160 7.2.1PhysicalModel .......................... 160 7.2.2TransmissionRangeandSensing/InterferenceRange ...... 160 7.3ProblemsandTheDesiredProtocolBehavior .............. 161 7.3.1HiddenandExposedTerminalProblem .............. 161 7.3.2LimitationsofNAVSetupProcedure ............... 162 7.3.3ReceiverBlockingProblem .................... 163 7.3.4Intra-FlowContention ....................... 164 7.3.5Inter-owContention ....................... 165 7.3.6TheDesiredProtocolBehavior .................. 165 7.3.7LimitationofIEEE802.11MACUsingSingleChannel ..... 166 7.4DUCHA:ANewDual-ChannelMACProtocol ............. 166 7.4.1ProtocolOverview ......................... 166 7.4.2BasicMessageExchange ..................... 167 7.4.3SolutionstotheAforementionedProblems ............ 169 7.4.4Remarksontheproposedprotocol ................ 171 7.5PerformanceEvaluation .......................... 172 7.5.1SimulationEnvironments ..................... 172 7.5.2SimpleScenarios ......................... 173 7.5.3RandomTopologyforOne-hopFlows .............. 176 7.5.4RandomTopologyforMultihopFlows .............. 178 7.6Conclusions ................................ 181 8ASINGLE-CHANNELSOLUTIONTOHIDDEN/EXPOSEDTERMINALPROBLEMSINWIRELESSADHOCNETWORKS ............ 183 8.1Introduction ................................ 183 8.2VariousRangesinWirelessMultihopAdHocNetworks ......... 188 8.3AddressingtheHidden/ExposedTerminalProblemswithShortBusyAdvertisementSignal .......................... 189 8.3.1BasicOperationsintheSBAProcedure .............. 190 viii

PAGE 9

............................ 191 8.3.3ParametersinSBAProcedure ................... 191 8.3.4PositionsofIDFSPeriodsintheDATAFrame .......... 193 8.3.5BusyAdvertisementSignal .................... 195 8.3.6PowerControlforShortBusyAdvertisement ........... 195 8.3.7StartandStopSBAProcedure .................. 196 8.3.8SynchronizationIssue ....................... 198 8.3.9AccumulativeAcknowledgement ................. 198 8.3.10CTSDominance .......................... 199 8.3.11CompatibilitywithLegacy802.11MACScheme ........ 199 8.4MaximizeSpatialReuseRatioandMinimizePowerConsumptionbyPowerControl ............................. 199 8.4.1PowerControlforBothDATAFrameandBusyAdvertisementinSBA-MAC .......................... 200 8.4.2PowerControlfortheApproachUsingALargeCarrierSensingRange .............................. 202 8.5PerformanceAnalysis ........................... 204 8.5.1SpatialReuseRatio ........................ 204 8.5.2ProtocolOverhead ......................... 204 8.5.3NumericalResults ......................... 206 8.6Conclusions ................................ 208 9ADISTRIBUTEDPACKETCONCATENATIONSCHEMEFORSENSORANDADHOCNETWORKS ......................... 211 9.1Introduction ................................ 211 9.2OperationsoftheIEEE802.11 ...................... 213 9.3AdaptivePacketConcatenation(APC)SchemeandPerformanceAnalysis 214 9.3.1BasicScheme ........................... 214 9.3.2PerformanceAnalysisoftheNetworkThroughputintheSingleHopCase ............................ 217 9.3.3PerformanceAnalysisoftheNetworkThroughputinaMultihopNetwork ............................. 221 9.4Conclusion ................................. 225 10IMPACTOFROUTINGMETRICSONPATHCAPACITYINMULTIRATEANDMULTIHOPWIRELESSADHOCNETWORKS ........... 226 10.1Introduction ................................ 226 10.2ImpactofMultirateCapabilityonPathSelectionInWirelessAdHocNetworks ................................ 231 10.2.1ReceiverSensitivityandSNRforMultipleRates ......... 231 10.2.2Tradeoffbetweentherateandthetransmissiondistance ..... 232 10.2.3CarrierSensingRange,InterferenceandSpatialReuse ...... 232 10.2.4EffectiveDataRateandProtocolOverhead ............ 233 ix

PAGE 10

............... 234 10.3.1LinkConictGraph ........................ 235 10.3.2UpperBoundofPathCapacityinSingleInterferenceModel ... 237 10.3.3ExactPathCapacityinSingleInterferenceModel ........ 240 10.3.4PathCapacityinMulti-InterferenceModelwithVariableLinkRate ............................... 242 10.3.5ExtendedtoMultiplePathsbetweenaSourceandItsDestina-tionorbetweenMultiplePairsofSourceandDestination ... 243 10.3.6Considerthepacketerrorrateovereachlinkinthelinkschedul-ingalgorithm .......................... 244 10.4PathSelectioninWirelessAdHocNetworks ............... 244 10.4.1OptimalPathSelection ...................... 245 10.4.2UsingRoutingMetricsinPathSelection ............. 246 10.5PerformanceEvaluation .......................... 247 10.5.1SimulationSetup ......................... 247 10.5.2ComparedwithOptimalRouting ................. 248 10.5.3PerformanceEvaluationofSixRoutingMetricsinaLargerTopol-ogy ............................... 249 10.5.4PathCapacityofaSingle-RateNetwork ............. 252 10.6Conclusions ................................ 253 11DISTRIBUTEDFLOWCONTROLANDMEDIUMACCESSCONTROLINMOBILEADHOCNETWORKS ....................... 255 11.1Introduction ................................ 255 11.2ImpactofMACLayerContentionsonTrafcFlows ........... 258 11.3OPET:OptimumPacketSchedulingforEachTrafcFlow ........ 261 11.3.1Overview ............................. 261 11.3.2Rule1:AssigningHighChannelAccessPrioritytoReceivers .. 261 11.3.3Rule2:Backward-PressureScheduling .............. 263 11.3.4Rule3:SourceSelf-ConstraintScheme .............. 268 11.3.5Rule4:RoundRobinScheduling ................. 270 11.4PerformanceEvaluation .......................... 271 11.4.1SimpleScenarios ......................... 272 11.4.2RandomTopology ......................... 273 11.4.3RandomTopologywithMobility ................. 276 11.4.4SimulationresultsforTCPtrafc ................. 277 11.4.5Notesontherelativebenetsofthefourtechniques ....... 279 11.5RelatedWorksandDiscussion ...................... 280 11.6Conclusions ................................ 282 12WCCP:IMPROVINGTRANSPORTLAYERPERFORMANCEINMULTI-HOPADHOCNETWORKSBYEXPLOITINGMACLAYERINFOR-MATION .................................... 283 12.1Introduction ................................ 283 x

PAGE 11

.................... 286 12.2.1TCPPerformanceDegradationDuetoCouplingofCongestionandMediumContention .................... 286 12.2.2OptimalCongestionWindowSizeforTCPandIdealSendingRate 288 12.2.3UnfairnessProblemDuetoMediumContention ......... 290 12.3WirelessCongestionControlProtocol(WCCP) ............. 292 12.3.1ChannelBusynessRatio:SignofCongestionandAvailableBand-width .............................. 292 12.3.2MeasurementofChannelBusynessRatioinMultihopAdHocNetworks ............................ 294 12.3.3Inter-nodeResourceAllocation .................. 295 12.3.4Intra-nodeResourceAllocation .................. 297 12.3.5End-to-EndRate-BasedCongestionControlScheme ....... 299 12.4PerformanceEvaluation .......................... 302 12.4.1ChainTopology .......................... 303 12.4.2RandomTopology ......................... 308 12.5Conclusions ................................ 308 13CONCLUSIONSANDFUTUREWORK .................... 310 13.1FairnessinMobileAdHocNetworks ................... 310 13.2QualityofServiceinMobileAdHocNetworks ............. 313 REFERENCES ..................................... 315 BIOGRAPHICALSKETCH .............................. 328 xi

PAGE 12

Table page 2IEEE802.11systemparameters ........................ 22 2Saturationvalueofcollisionprobability .................... 22 3QoSrequirementsformultimediaservices .................. 36 3IEEE802.11systemparameters ........................ 42 4IEEE802.11systemparameters ........................ 71 4Themean,standarddeviation(SD),and97'th,99'th,99.9'thpercentilede-lays(inseconds)forvoiceandvideointheinfrastructuremode. ...... 83 4Themean,standarddeviation(SD),and97'th,99'th,99.9'thpercentilede-lays(inseconds)forvoiceandvideointheadhocmode. ......... 85 6Signal-to-noiseratioandreceiversensitivity ................. 131 7Defaultvaluesinthesimulations ........................ 172 9IEEE802.11systemparameters ........................ 220 10Signal-to-noiseratioandreceiversensitivity ................. 232 10Runtimeofdifferentroutingalgorithms .................... 253 12SimulationresultsforTCPandUDPows .................. 289 12PerformanceofWCCPandTCPinchaintopologyofFig. 12 (a) ..... 303 xii

PAGE 13

Figure page 2RTS/CTSmechanismandbasicaccessmechanismofIEEE802.11 .... 11 2Generalizedstatetransitiondiagramofoneexample ............ 15 2Generalizedstatetransitiondiagramfortransmissionprocess ....... 19 2ProbabilitydistributionofMAClayerservicetime ............. 21 2PDFofservicetime ............................. 23 2Meanofservicetime ............................. 23 2Queuecharacteristics ............................. 28 2MAClayerpacketservicetime ....................... 30 2ComparisonsbetweenM/G/1/K,M/M/1/Kmodelsandsimulation ..... 31 2Averagewaitingtimeinnon-saturatedstatus ................ 32 2AverageMAClayerservicetime ...................... 33 3Channelbusynessratioandutilization .................... 41 3CollisionprobabilityandmaximumnormalizedthroughputwithRTS/CTSandpayloadsizeof8000bits ....................... 45 3ImpactofpayloadsizeandtheRTS/CTSmechanism ............ 47 3Meanandstandarddeviationofservicetime ................ 49 3Packetdelay ................................. 54 3Simulationresultswhenpayloadsize=8000bits .............. 57 3Simulationresultswhenn=50andpayloadsize=8000bits ......... 58 3Simulationresultswhenn=50andpayloadsize=8000bits ......... 60 4Channelbusynessratioandutilization .................... 70 4Simulationresultswhennumberofnodesequals50andRTS/CTSmech-anismisused ............................... 71 xiii

PAGE 14

.................... 82 4End-to-enddelayofallvoiceandvideopacketsininfrastructuremode .. 83 4Adhocmode:thenumberofreal-timeandTCPowsincreasesovertime.Channelrateis2Mbps. .......................... 84 4End-to-enddelayofallvoiceandvideopacketsinadhocmode ...... 85 4Channelutilizationinadhocmode ..................... 86 5Maximumandsaturatedthroughputwithdifferentnumberofnodes(RTS/CTSisused,packetlength=1000bytes,channelrate=11Mbps) ....... 94 5Convergencespeedofmultiplicative-increasephase(packetlength=1000bytes,channelrate=11Mbps) .......................... 105 5ConvergencespeedofAIMDphaseswhen=0:5 109 5ImpactofpayloadsizeLandnumberofnodesnontheoptimalthresholdforchannelbusynessratiobrth 111 5FairnessconvergencewithRTS/CTS:onegreedynodejoinsthenetworkevery10seconds(packetlength=1000bytes,eachpointisaveragedover1second) ............................... 113 5Max-minfairnessunderdifferenttrafcrates(packetlength=1000bytes) 114 5DRA:fairnesswithmultiplechannelbitrates(RTS/CTSisused) ..... 115 5802.11:fairnesswithmultiplechannelbitrates(RTS/CTSisused) .... 115 5Throughput,MACdelayandcollisionprobabilitywithRTS/CTS ..... 117 5QoSperformanceinDRA .......................... 118 5QoSperformancein802.11 ......................... 118 6Interferencemodel .............................. 128 6CarriersensingthresholdwithShanonCapacity .............. 130 6CarriersensingthresholdwithdifferentSINR ................ 131 6Carriersensingthresholdwithdiscretechannelratesof802.11 ....... 132 6Tradeoffbetweenexposedterminalproblemandhiddenterminalproblem 134 6LargecarriersensingrangewithcarriersensingstrategyIIforCTS/ACK 139 6Multiplecarriersensingthresholdsmayresultincollisions ......... 141 xiv

PAGE 15

......................... 144 6Maximumend-to-endthroughputfordifferenthopdistance ........ 145 6Spatialreuseratioformultihopows(a)atworstcase,(b)inasinglechaintopologywithonewaytrafc ....................... 147 6Optimumcarriersensingthresholdforone-hopows ............ 152 6Optimumcarriersensingthresholdformulti-hopows ........... 152 7Asimplescenariotoillustratetheproblems ................. 162 7Chaintopology ................................ 165 7Proposedprotocol .............................. 167 7Onesimpletopology ............................. 173 7Simulationresultsforthesimpletopology ................. 174 7End-to-endthroughputforthe9-nodechaintopology ............ 177 7Simulationresultsforrandomone-hopowswithdifferentminimumonehopdistance ................................ 177 7Simulationresultsformultihopowsinrandomtopology ......... 179 8Hiddenterminalproblem .......................... 184 8CarriersensingrangeandinterferencerangeinLCSandSBA-MAC .... 185 8Four-wayhandshakewithbusyadvertisementsignals ............ 190 8PositionsofIDFSperiodsintheDATAframe ................ 193 8PowercontrolinSBA-MAC ......................... 200 8Occupiedareaforatransmissionnormalizedoverthecommunicationra-dius(PC:powercontrolforDATAframes) ................ 207 8Occupiedareaforatransmissionnormalizedoverthecommunicationra-diuswhendh=dt 207 8Channeltimeforatransmittedpacket .................... 208 8Channeltimeforatransmittedpacket .................... 209 8PerformancegainofSBA-MACcomparedtotheapproachusingalargecarriersensingrangeandtheFAMAscheme ............... 209 9RTS/CTSmechanismandbasicaccessmechanismofIEEE802.11 .... 214 xv

PAGE 16

................................ 215 9Thesuperpacketstructure .......................... 215 9Throughputwhenchannelrateis1Mbps,Lth=2346bytesandRTS/CTSmechanismisused. ............................ 220 9Throughputwhenchannelrateis1,2,5.5and11MbpsandRTS/CTSmechanismisused. ............................ 221 9Chaintopology. ............................... 222 9Maximumend-to-endthroughputofamultihopow. ............ 224 9Maximumend-to-endthroughputofamultihopow. ............ 224 10PathsbetweenthesourceSandthedestinationD 230 10Ave-linkchaintopologyanditslinkConictgraph ............ 235 10Apathwithanoddcycleinthelinkconictgraph ............. 239 10Pathcapacityfordifferentroutingalgorithms ................ 249 10Pathcapacityfordifferentroutingalgorithms ................ 250 10Pathlengthsfordifferentroutingalgorithms ................ 251 10Source-destinationdistance ......................... 251 10Pathcapacitysolvingtime .......................... 252 10Pathcapacityforasingleratenetwork ................... 254 11Chaintopologyandcrosstopology ..................... 259 11TCPperformanceina9-nodechaintopology ................ 260 11Optimumpacketschedulingforchaintopology ............... 263 11ThepacketformatofRTSMandCTSR ................... 266 11Thealgorithmsofbackward-pressurescheme ................ 267 11Messagesequenceforpackettransmission ................. 268 11Thepacketschedulingforresolvingcongestion ............... 269 11Simulationresultsforthe9-nodechaintopology(Fig. 11 )andcrosstopology(Fig. 11(b) ) .......................... 272 11Simulationresultsfortherandomtopology ................. 274 xvi

PAGE 17

......... 277 11SimulationresultsfortheTCPtrafc .................... 278 11Gridtopologywith16TCPows ...................... 279 12Chaintopologywith9nodes.Smallcirclesdenotethetransmissionrange,andthelargecirclesdenotethesensingrange ............... 286 12Simulationresultsfor9-nodechaintopology ................ 287 12Nine-nodechaintopologywithdifferenttrafcdistribution ......... 291 12Therelationshipbetweenchannelbusynessratioandothermetrics ..... 293 12Ratecontrolmechanism ........................... 300 12Simulationresultsforthenine-nodechaintopologywithoneow ..... 304 12PerformanceofscenarioFig. 12 (b) .................... 305 12PerformanceofscenarioFig. 12 (c) .................... 306 12Simulationresultsforrandomtopologywithprecomputedpaths:(a)min-imumowthroughputin20runs,(b)minimumowthroughputaver-agedover20runs,(c)maximumowthroughputaveragedover20runs,(d)ratioofaveragedmaximumowthroughputtoaveragedminimumowthroughput. .............................. 307 12Simulationresultsaveragedover20runsintherandomtopology:(1)ag-gregatethroughput(Mbps),(2)fairnessindex,(3)end-to-enddelay(s). 308 13Anoriginaltopologyanditsowcontentiongraph ............. 311 xvii

PAGE 18

ThisPh.D.dissertationfocusesondesignandanalysisofefcientnetworkingproto-colsinwirelesslocalareanetworksandadhocnetworks.KnownasWI-FItechnology,wirelesslocalareanetworkshavebecomeverypopulartodayasaneasywayofwire-lessaccesstotheInternet.Wirelessadhocnetworksalsondalotofapplicationswhichneedwirelessaccessorrequirealowcostoranimmediatedeploymentofnetworkedsys-tems,likebattleeldcommunications,publicsafetynetworks,disasterrescues,andwire-lessmetropolitanareanetworks.However,itisaverychallengingtasktodesignefcientnetworkingprotocolstoprovidequalityofservice(QoS)andreliabilityinthesenetworks.Comparedtowirednetworks,inwirelessnetworkslinksarenotindependentanymore;bandwidth,powerandprocessingabilityarelimited;channelerrorshappenfrequently;networktopologyissubjecttoconstantchange;andthenetworkisoftenself-organizedanddistributed.Thesechallengesleadtoclosecouplingamongvariouslayersintheproto-colstackandacompletedifferentmediumaccesscontrol(MAC)layer,andhencecallforacross-layerdesignbetweentheMAClayerandotherlayers. Thedissertationconductsathoroughtheoreticalstudyonacontention-basedMACstandard,IEEE802.11,andinvestigatestheclosecouplingbetweenvariousprotocollayers xviii

PAGE 19

xix

PAGE 20

However,therearealotofchallengesforthenetworkingprotocolstoworkefcientlyinWLANsandMANETs.Unlikewirednetworks,someuniquecharacteristicsofWLANsandMANETsseriouslydeteriorateperformanceofthenetworkingprotocols.Thesechar-acteristicsincludethetime-varyingchannelsduetopathloss,fadingandinterference,thevulnerablesharedmediaaccessduetorandomaccesscollisionandthelimitedbatteryen-ergy.InMANETs,thenetworktopologymayexperiencecontinuouschangeandcausefre-quentroutebreakagesandre-routingactivity.AndMANETsbynatureareself-organized,self-controlledanddistributed.Inotherwords,thereisnocentralizedcontrollerthathasperfectknowledgeofallthenodesinthenetwork.Instead,eachnodecanonlyhaveincom-pleteorsometimesskewedviewofthenetwork.Asaresult,ithastomakedecisionswithimperfectinformation.DuetoallthesehurdlesposedbyWLANsandMANETs,simple,efcient,fair,andenergy-efcientnetworkingprotocols,whilehighlydesirable,arenotaneasytask. 1

PAGE 21

Thesechallengescallforthecross-layerdesignofthenetworkingprotocolsinWLANsandMANETs.Forexample,byschedulingthenodewithgoodchannelqualitytoaccessthechannel,mediumaccesscontrol(MAC)protocolscanachievehigherthroughput.ThetraditionalcongestioncontrolprotocolfortheInternet,TCP,takesanypacketlossasacongestionsign.However,packetlosscanbeattributedtopoorchannelqualityorroutefailureduetomobility.ItcanachievebetterperformanceiftheTCPsourcecandiffer-entiatethedifferentreasonsofpacketlossesbyobtaininginformationfromtheroutingprotocolsandthephysicalandMAClayers.Routingprotocolscanalsoavoidunnecessaryre-routingmessagesiftheycandistinguishthepacketlossesformediumcollisioninsteadofmobility.Astoqualityofservice(QoS)andfairness,theyseemtobeformidabletasksconsideringtheunreliablephysicalchannel,mediumcollisionsanddynamicallychangingnetworktopologyandtrafcload.Cross-layerdesignseemsamusttoprovidenode-basedandow-basedfairnessandend-to-endQoSguarantee. 2 .Wepro-poseanewmodelusingthesignaltransferfunctionofgeneralizedstatetransferdiagramtocharacterizetheprobabilitygenerationfunctionofthemediumaccessdelay.Withtheprob-abilitydistributionofmediumaccessdelayandqueueingtheory,mostoftheperformancemetrics,suchasthroughput,delay,packetlossrateandvariousqueuecharacteristics,canbeanalyzedfortheWLANs.Ourresultsshowthatatthenon-saturatedstate(i.e.,eachnodedoesnotcontendforthechannelallthetimeandthetotaltrafcratedoesnotexceedthenetworkcapacity),theperformanceisdependentonthetotaltrafcandalmostindiffer-enttothenumberoftransmittingstations.Atthesaturatedstate(i.e.,eachnodehasenoughtrafctokeepcontendingthesharedwirelesschannel),thenumberoftransmittingstationsaffectstheperformancemoresignicantly.

PAGE 22

InChapter 3 ,wefurtherderivethemaximumthroughputoftheIEEE802.11DCFprotocolandaccurateestimatesfordelayanddelayvariationsinwirelessLANsbasedonourworkinChapter 2 .Weshowthat,bycontrollingthetotaltrafcrate,theoriginal802.11DCFprotocolcansupportstrictQoSrequirements,suchasthoserequiredbyvoiceoverIPorstreamingvideo,andatthesametime,achieveahighchannelutilization.TheresultisasignicantdeparturefrommostrecentworkswhichonlysupportservicedifferentiationinsteadofQoSguarantee. ThestudiesinChapter 3 alsosuggestagoodmetricchannelbusynessratiotorepresentthenetworkstatus,suchasthroughput,mediumaccessdelayandcollisionprobability.Justasthenameimplies,channelbusynessratioisaratioofthetimeintervalswhenthechannelisbusyduetosuccessfultransmissionsandcollisionstothetotaltime.BasedonthephysicalcarriersensingandvirtualcarriersensingmechanismsoftheIEEE802.11standard,thismetricisveryeasytomeasureandonlyrequiresafewsimplecalculationsattheMAClayer.HenceitcanbeusedtofacilitatetheregulationoftotalinputtrafctosupportQoS. InChapter 4 ,weproposeacalladmissionandratecontrolschemetosupportQoSguaranteeinWirelessLANs.BasedonthechannelbusynessratioobtainedattheMAClayer,thecalladmissioncontrolalgorithmisusedtoregulatetheadmissionofreal-timeandstreamingtrafcandtheratecontrolalgorithmtocontrolthetransmissionrateofbestefforttrafc.Asaresult,thereal-timeorstreamingtrafcissupportedwithstatisticalQoSguarantee,andthebestefforttrafccanfullyutilizetheresidualchannelcapacityleftbythereal-timeandstreamingtrafc. InChapter 5 ,wefurtherdeveloptheschemeinChapter 4 intoacomprehensivepro-tocol.Fairnessisamajorfocusofthischapter.Weproposeanovelthree-phasecontrolmechanismtofairlyandefcientlyutilizenetworkresourceandguaranteeashortmediumaccessdelay.Theprotocolalsointegratesthethree-phasecontrolmechanismwithacall

PAGE 23

admissioncontrolschemeandapacketconcatenationschemeintoasingleuniedframe-worktobettersupportQoSandmultiplechannelratesbesidestheefciencyandfairness. AfterweexaminetheperformanceofwirelessLANsandproposeaschemetosup-portQoSaswellashighefciency,wearewonderingwhetherthesetechniquescanbeappliedtomultihopcase,i.e.,MANETs.However,MANETsaremuchmorecomplicatedthanwirelessLANs.Therearealotofnewchallenges,suchastheinfamoushiddenandexposedterminalproblems.Beforewecomeupwithanydesigns,wemustrstunderstandthoroughlywhattheproblemsareandhowtheyimpactthenetworkperformance. InChapter 6 ,westudytheimpactofphysicalcarriersensingandvirtualcarriersens-ingmechanismsonthesystemperformanceofMANETs.Atheoreticalmodelisdevelopedtoanalyzetheoptimalcarriersensingrangetomaximizethesystemthroughputwhenmul-tiplediscretechannelratescoexistinthenetwork.Wealsostudyhowtoutilizethemulti-ratecapabilityoftheIEEE802.11standard,andwhichneighborandchannelrateshouldbeusedforeachhoptransmission.Anovelroutingmetric,bandwidthdistanceproduct,isproposedtoperformthistaskanditcangreatlyincreasethesystemthroughput. InChapter 7 ,werststudythevariousproblemsofthemediumaccesscontroliftheIEEE802.11DCFprotocolisused,suchasthehiddenandexposedterminalproblems,receiverblockingproblemandintra-owandinter-owcontentionproblems.ThestudiesshowthattheseproblemsnotonlyimpacttheefciencyoftheMACprotocolbutalsoim-pactthehigherlayers'performance,suchasunnecessaryre-routingactivitiesduetofalseroutefailuresandunfairnessamongmultipleows.Motivatedbytheanalysisoftheseproblems,weproposeanewdual-channelMACprotocol.ThenewMACprotocolusesanout-of-bandbusytoneandtwocommunicationchannels,oneforcontrolframesandtheotherfordataframes.Thenewlydesignedmessageexchangesequenceprovidesacom-prehensivesolutiontoalltheaforementionedproblems.Extendedsimulationsdemonstratethatourschemeprovidesamuchmorestablelinklayer,greatlyimprovesthespatialreuse,andworkswellinreducingthepacketcollisions.Itimprovesthethroughputby8%to28%

PAGE 24

forone-hopowsandby25timesformultihopowsunderheavytrafccomparingtotheIEEE802.11MACprotocol. However,sometimesweonlyhaveonesinglechannelandonesingletransceiver.InthiscaseweneedtodevelopanewefcientMACprotocolotherthanDUCHAtoaddressthoseproblems.Therefore,inChapter 8 ,weproposeacompletesinglechannelsolutiontoaddressbothhiddenandexposedterminalproblems.ThenewsolutioninsertsdummybitsintheDATAframeandallowsthereceivertotransmitshortbusyadvertisementsduringthetransmissiontimeofthedummybitstonotifythehiddenterminaloftheongoingtransmis-sion.BecausethetransmissionofDATAframeisprotectedbytheshortbusyadvertisementsignals,weareabletosignicantlyreducethecarriersensingrangetoincreasethespatialreuseratio,whichnoticeablymitigatetheexposedterminalproblem.Wealsodemonstratethatpowercontrolinthesolutioncanfurtherremarkablyimprovethesystemperformance. InChapter 9 ,westudyhowthephysicallayerinformationcanbeusedattheMAClayertoimprovethesystemperformance.Weproposeanewadaptivepacketconcatenation(APC)schemeanddemonstratethatAPCcanimprovethesystemthroughputbyseveraltimesinbothWLANsandMANETs. InChapter 10 ,wefocusontheimpactofroutingmetricsonthethroughputofse-lectedpathsinMANETs.BecauseMAClayerandPhysicallayerhaveagreatimpactontheroutingalgorithm,consideringthefeaturesofthesetwolayersisamustinagoodrout-ingalgorithm.Werstperformacomprehensivestudyontheimpactofmultiplerates,interferenceandpacketlossratetogetheronthemaximumend-to-endthroughputorpathcapacity.Atheoreticalmodelisderivedtostudythepathcapacityorthemaximumend-to-endthroughputofselectedpathswithconsiderationofallthosefactors.WealsoproposeanewroutingmetriccalledinterferencecliquetransmissiontimetoefcientlyutilizetheinformationatphysicalandMAClayerstondgoodpaths.Basedontheproposedthe-oreticalmodel,weevaluatethecapabilityofvariousroutingmetricsincludinghopcount,expectedtransmissiontimes,end-to-endtransmissiondelayormediumtime,linkrate,

PAGE 25

bandwidthdistanceproduct,interferencecliquetransmissiontime,tondapathwithhighthroughput.Theresultsshowthatinterferencecliquetransmissiontimeisabetterroutingmetricthanalltheothers. InChapter 11 ,bycarefullystudyingtheintra-owandinter-owcontentionproblems,wendthatnetworkcongestioniscloselycoupledwiththemediumaccesscontentions.ThenweproposeaframeworkofdistributedowcontrolandmediumaccesstomitigatetheMAClayercontentions,overcomethecongestionandincreasethethroughputfortrafcowsacrosssharedchannelenvironments.Thekeyideaisbasedontheobservationthat,intheIEEE802.11MACprotocol,themaximumthroughputforastandardchaintopologyis1/4ofthechannelbandwidthanditsoptimumpacketschedulingistoallowsimultaneoustransmissionsatnodeswhicharefourhopsaway.Theproposedfullydistributedschemegeneralizesthisoptimumschedulingtoanytrafcowwhichmayencounterintra-owandinter-owcontentions.OurschemehasbeenshowntoperformbetterandachievehigherthroughputatlighttoheavytrafcloadcomparingtothatwhentheoriginalIEEE802.11MACprotocolisused.Moreover,ourschemealsoachievesmuchbetterandmorestableperformanceintermsofdelay,fairnessandscalabilitywithlowandstablecontroloverhead. TheproposedschemeinChapter 11 providesagoodsolutionofcongestioncontrolatthenetworkanddatalinklayers.However,tosupportend-to-endreliabilityrequiredbyvariousservices,suchaswebtrafcandemails,end-to-endowandcongestioncontrolisalsonecessary.Chapter 12 studiestheclosecouplingbetweenTCPtrafcandmediumcon-tentionandndsthattheTCPsourcesareverygreedyleadingtoseverenetworkcongestionandmediumcollisions.Andthewindowbasedcongestioncontrolalgorithmbecomestoocoarseinitsgranularity,causingthroughputinstabilityandexcessivelylongdelay.Basedonthenoveluseofchannelbusynessratio,whichweshowinChapter 3 isanaccuratesign

PAGE 26

ofthenetworkutilizationandcongestionstatus,anewend-to-endcongestioncontrolpro-tocolhasbeenproposedtoefcientlyandfairlysupportthetransportserviceinmultihopadhocnetworks. Finally,Chapter 13 discussessomefutureresearchissuesincludingthefairnessandQoSsupportinMANETs.

PAGE 27

IEEE802.11MACprotocolisthedefactostandardforwirelessLANs,andhasalsobeenimplementedinmanynetworksimulationpackagesforwirelessmulti-hopadhocnetworks.However,itiswellknownthat,asthenumberofactivestationsincreases,theperformanceofIEEE802.11MACintermsofdelayandthroughputdegradesdramati-cally,especiallywheneachstation'sloadapproachestoitssaturationstate.Toexploretheinherentproblemsinthisprotocol,itisimportanttocharacterizetheprobabilitydis-tributionofthepacketservicetimeattheMAClayer.Inthischapter,bymodelingtheexponentialbackoffprocessasaMarkovchain,wecanusethesignaltransferfunctionofthegeneralizedstatetransitiondiagramtoderiveanapproximateprobabilitydistributionoftheMAClayerservicetime.WethenpresentthediscreteprobabilitydistributionforMAClayerpacketservicetime,whichisshowntoaccuratelymatchthesimulationdatafromnetworksimulations.BasedontheprobabilitymodelfortheMAClayerservicetime,wecananalyzeafewperformancemetricsofthewirelessLANandgivebetterexplanationtotheperformancedegradationindelayandthroughputatvarioustrafcloads.Furthermore,wedemonstratethattheexponentialdistributionisagoodapproximationmodelfortheMAClayerservicetimeforthequeueinganalysis,andthepresentedqueueingmodelscanaccuratelymatchthesimulationdataobtainedfromns-2whenthearrivalprocessatMAClayerisPoissonian. 8

PAGE 28

However,therearemanyproblemsencounteredinthehigherprotocollayersinIEEE802.11wirelessnetworks.Ithasbeenobservedthatthepacketdelayincreasesdramati-callywhenthenumberofactivestationsincreases.PacketsmaybedroppedeitherduetothebufferoveroworbecauseofseriousMAClayercontentions.SuchpacketlossesmayaffecthighlayernetworkingschemessuchastheTCPcongestioncontrolandnetworkingroutingmaintenance.Theroutingsimulations[ 19 108 ]overmobileadhocnetworksindi-catethatnetworkcapacityispoorlyutilizedintermsofthroughputandpacketdelaywhentheIEEE802.11MACprotocolisintegratedwithroutingalgorithms.TCPinthewirelessadhocnetworksisunstableandhaspoorthroughputduetoTCP'sinabilitytorecognizethedifferencebetweenthelinkfailureandthecongestion.Besides,oneTCPconnectionfromone-hopneighborsmaycapturetheentirebandwidth,leadingtotheone-hopunfairnessproblem[ 64 140 46 110 ]. PerformanceanalysisfortheIEEE802.11MACprotocolcouldhelptodiscovertheinherentcauseoftheaboveproblemsandmaysuggestpossiblesolutions.Manypapersonthistopichavebeenpublished[ 20 22 15 44 58 134 85 ].Cali[ 20 22 ]derivedtheprotocolcapacityoftheIEEE802.11MACprotocolandpresentedanadaptiveback-offmechanismtoreplacetheexponentialbackoffmechanism.Bianchi[ 15 ]proposedaMarkovchainmodelforthebinaryexponentialbackoffproceduretoanalyzeandcomputetheIEEE802.11DCFsaturatedthroughput.Allofthesepapersassumethesaturatedsce-nariowhereallstationsalwayshavedatatotransmit.BasedonthesaturatedthroughputinBianchi'smodel,FohandZuckermanpresentedtheanalysisofthemeanpacketdelayatdifferentthroughputforIEEE802.11MAC[ 44 ].Hadzi-VelkovandSpasenovskialsogaveananalysisforthethroughputandmeanpacketdelayinthesaturatedcasebyincor-poratingframe-errorrates[ 58 ].KimandHou[ 85 ]analyzedtheprotocolcapacityofIEEE802.11MACwiththeassumptionthatthenumberofactivestationshavingpacketsreadyfortransmissionislarge.

PAGE 29

Totheauthors'bestknowledge,thereisnocomprehensivestudyonthequeuedy-namicsoftheIEEE802.11wirelessLANs.Thedelayanalysisislimitedtothederivationofmeanvaluewhilethehighermomentsandtheprobabilitydistributionfunctionofthedelayareuntouched.Mostofthecurrentpapersfocusedontheperformanceanalysisinsaturatedtrafcscenariosandthecomprehensiveperformancestudyundernon-saturatedtrafcsituationsisstillopen. Inthischapter,toaddresstheaboveissues,werstcharacterizetheprobabilitydis-tributionoftheMAClayerpacketservicetime(i.e.,thetimeintervalbetweenthetimeinstantapacketstartstocontendfortransmissionandthetimeinstantthatthepacketeitherisacknowledgedforcorrectreceptionbytheintendedreceiverorisdropped).BasedontheprobabilitydistributionmodeloftheMAClayerpacketservicetime,wethenstudythequeueingperformanceofthewirelessLANsatdifferenttrafcloadbasedontheIEEE802.11MACprotocol.Then,weevaluatetheaccuracyoftheexponentialprobabilitydistri-butionmodelfortheMAClayerservicetimeinqueueinganalysisthroughbothanalyticalapproachandsimulations. 2.2.1DistributedCoordinationFunction(DCF) 68 ].IntheDCFprotocol,astationshallensurethatthemediumisidlebeforeattemptingtotransmit.Itselectsarandombackoffintervallessthanorequaltothecurrentcontentionwindow(CW)sizebasedontheuniformdistribution,andthendecreasesthebackofftimerbyoneateachtimeslotwhenthemediumisidle(maywaitforDIFSfollowedasuccessfultransmissionorEIFSfollowedacollision).Ifthemediumisdeterminedtobebusy,thestationwillsuspenditsbackofftimeruntiltheendofthecurrenttransmission.Transmissionshallcommencewheneverthebackofftimerreacheszero.Whentherearecollisionsduringthetransmissionorwhenthetransmissionfails,thestationinvokesthebackoffprocedure.Tobeginthe

PAGE 30

backoffprocedure,thecontentionwindowsizeCW,whichtakesaninitialvalueofCWmin,doublesitsvaluebeforeitreachesamaximumupperlimitCWmax,andremainsthevalueCWmaxwhenitisreacheduntilitisreset.Then,thestationsetsitsbackofftimertoarandomnumberuniformlydistributedovertheinterval[0,CW)andattemptstoretransmitwhenthebackofftimerreacheszeroagain.Ifthemaximumtransmissionfailurelimitisreached,theretransmissionshallstop,CWshallberesettoCWmin,andthepacketshallbediscarded[ 68 ].TheRTS/CTSmechanismsandbasicaccessmechanismofIEEE802.11areshowninFig. 2 Figure2: RTS/CTSmechanismandbasicaccessmechanismofIEEE802.11 TheservicetimeofthequeueingsystemistheMAClayerpacketservicetimedenedinSection 2.1 .TheIEEE802.11MACadoptsthebinaryexponentialbackoffmechanismforthetransmissionofeachpacket,whichmaycollidewithsomeothertransmissionsin

PAGE 31

theairateachtransmissionattempt.Andthecollisionprobabilitypcisdeterminedbytheprobabilitythatthereisatleastoneofotherstationswhichwilltransmitatthesamebackofftimeslotwhentheconsideredstationattemptstransmission.Weassumethatthisprobabil-itydoesnotchangeandisindependentduringthetransmissionofeachpacketregardlessofthenumberofretransmissionsuffered.Forthesaturatedcase,thisapproximationhasbeenusedbyBianchi[ 15 ]toderivethesaturatedthroughput.Andforthenon-saturatedcase,thecollisionprobabilitybecomesmorecomplex.Itdependsonthenumberofstationswithpacketsreadyfortransmissionandthebackoffstatesofthesestations.Betweentwotrans-missionattemptsattheconsideredstation,otherstationsmaycompleteseveralsuccessfultransmissionsand/orencounterseveralcollisions,andtheremaybenewpacketarrivalsatstationsnomatterwhethertheyarepreviouslycontendingfortransmissionornot.Intu-itively,thisapproximationbecomesmoreaccuratewhenthenumberofstationsgetslargerforbothsaturatedandnon-saturatedcase.Forsimplicity,weusethesameapproximationforbothcasesandarguethatthecollisionprobabilitydoesnotchangesignicantlyaslongastheinputtrafcratefromhigherlayerateachstationarestillthesameduringtheser-viceforeachpacket.ThenwecouldmodelthebinaryexponentialbackoffmechanismasaMarkovchainandmakepossiblethederivationoftheprobabilitydistributionofservicetimeinthenextsection.Laterinthischapter,wewillshowthattheanalyticalresultsfromthisapproximationareconsistentwiththesimulationresultsverywellatthenon-saturatedcase. 2.3.1MACLayerServiceTime 2.2 ,therearethreebasicprocesseswhentheMAClayertrans-mitsapacket:thedecrementprocessofthebackofftimer,thesuccessfulpackettransmis-sionprocessthattakesatimeperiodofTsucandthepacketcollisionprocessthattakesatimeperiodofTcol.Here,Tsucistherandomvariablerepresentingtheperiodthatthe

PAGE 32

mediumissensedbusybecauseofasuccessfultransmission,andTcolistherandomvari-ablerepresentingtheperiodthatthemediumissensedbusybyeachstationduetocolli-sions. TheMAClayerservicetimeisthetimeintervalfromthetimeinstantthatapacketbecomestheheadofthequeueandstartstocontendfortransmissiontothetimeinstantthateitherthepacketisacknowledgedforasuccessfultransmissionorthepacketisdropped.Thistimeisimportantwhenweexaminetheperformanceofhigherprotocollayers.Appar-ently,thedistributionoftheMAClayerservicetimeisadiscreteprobabilitydistributionbecausethesmallesttimeunitofthebackofftimerisatimeslot.TsucandTcoldependonthetransmissionrate,thelengthofthepacketandtheoverhead(withadiscreteunit,i.e.,bit),andthespecictransmissionscheme(thebasicaccessDATA/ACKschemeortheRTS/CTSscheme)[ 15 68 ]. andcompletelycharacterizesthediscreteprobabilitydistributionofTS,andhasafewimportantpropertiesasfollows: @ZPTS(Z)Z=1=P0TS(1)VAR[X]=P00TS(1)+P0TS(1)fP0TS(1)g2(2.2) wheretheprimeindicatesthederivative.

PAGE 33

ToderivethePGFoftheMAClayerservicetime,wewillmodelthetransmissionprocessofeachpacketasaMarkovchaininthefollowingsubsections.HerewerstdiscusshowtodrivethePGFoftheservicetimefromtheMarkovchain. Thestatewhenthepacketleavesthemobilestation,i.e.,beingsuccessfullytransmit-tedordropped,istheabsorptionstateoftheMarkovchainforthebackoffmechanism.ToobtaintheaveragetransitiontimetotheabsorptionstateoftheMarkovchain,wecanusethematrixgeometricapproach.However,inthecaseofMarkovChainforTSwithvarioustransitiontimesondifferentbranches,itrequiresanewmatrixformulationtoaccommo-datedifferenttransitiontimes,anditssolutionalwaysaccompaniesextraneouscomplicatedcomputations[ 30 ].Here,weapplythegeneralizedstatetransitiondiagram,fromwhichwecaneasilyderivethePGFofTSandobtainarbitrarynthmomentofTS. Inthegeneralizedstatetransitiondiagram,wemarkthetransitiontimeoneachbranchalongwiththetransitionprobabilityinthestatetransitiondiagram(theMarkovchain).Thetransitiontime,whichisthedurationforthestatetransitiontotakeplace,isexpressedasanexponentofZvariableineachbranch.Thus,theprobabilitygeneratingfunctionoftotaltransitiontimecanbeobtainedfromthesignaltransferfunctionofthegeneralizedstatetransitiondiagramusingthewell-knownMasonformula[ 30 112 ]. ToillustratehowthegeneralizedMarkovchainmodelworks,weshowonesimpleexampleforaMACmechanismthatallowsinniteretransmissionsforeachpacketwithoutanybackoffmechanisms.IftherandomvariableFisdenedasthedurationoftimetakenforastatetransitionfromthestatetoinFig. 2 ,itsPGFissimplythesignaltransferfunctionofthestatetransition.InFig. 2 ,pisthecollisionprobability,1pisthesuccessfullytransmittedprobability,1isthecollisiontime,and2isthesuccessfultransmissiontime.SothePGFofrandomvariableFis

PAGE 34

ThissatisesEquation 2.2 ,thatis,PF(1)=1anditsmeantransitiontimeis Figure2: Generalizedstatetransitiondiagramofoneexample Ontheotherhand,wecaneasilyobtaintheaveragecollision/retransmissiontimesNC,i.e.,p=(1p).ThustheaveragetransitiontimecanbedirectlyobtainedasNC1+2,whichisthesameasEquation 2.4 2 ,theperiodofsuc-cessfultransmissionTsucequalsto AndtheperiodofcollisionTcolequalsto

PAGE 35

InthecasethatthelengthofDATAisaxedvaluelD,itsPGFSt(Z)equals Ifthebasicschemeisadopted,Tcolisdeterminedbythelongestoneofthecollidedpackets.Whentheprobabilityofthreeormorepacketssimultaneouslycollidingisne-glected,itsprobabilitydistributioncanbeapproximatedbythefollowingequation, whereli(i=1;2)isthepacketlengthoftheithcollidedpacket.Thuswecouldobtainthat (lmaxlmin+1)2lmaxXi=lmin(2i2lmin+1)Zi(2.11) forthecasethatthelengthofDATAhasauniformdistributionin[lmin,lmax],or forthecasethatthelengthofDATAisaxedvaluelD. Asmentionedinsection 2.2 ,pcistheprobabilityofacollisionseenbyapacketbeingtransmittedonthemedium.AssumingthattherearenstationsinthewirelessLANweareconsideringandpacketarrivalprocessesatallthestationsareindependentandidenticallydistributed,weobservethatpcisalsotheprobabilitythatthereisatleastone

PAGE 36

packettransmissioninthemediumamongother(n-1)stationsintheinterferencerangeofthestationunderconsideration.Thisyields wherep0istheprobabilitythattherearenopacketsreadytotransmitattheMAClayerinthewirelessstationunderconsideration,andisthepackettransmissionprobabilitythatthestationtransmitsinarandomlychosenslottimegiventhatthestationhaspacketstotransmit. LetPsucbetheprobabilitythatthereisonesuccessfultransmissionamongother(n-1)stationsintheconsideredslottimegiventhatthecurrentstationdoesnottransmit.Then, ThenpcPsucistheprobabilitythattherearecollisionsamongother(n-1)stations(orneighbors). Thus,thebackofftimerhastheprobabilityof1-pctodecrementby1afteranemptyslottime,theprobabilityPsuctostayattheoriginalstateafterTsuc,andtheprobabilityofpcPsuctostayattheoriginalstateafterTcol.SothedecrementprocessofbackofftimerisaMarkovprocess.Thesignaltransferfunctionofitsgeneralizedstatetransitiondiagramis Fromaboveformula,weobservethatHd(Z)isafunctionofpc,thenumberofstationsnandthedummyvariableZ.

PAGE 37

untilthemaximumretransmissionlimitisreachedafterTcol.SincethedecrementprocessofbackofftimerisaMarkovprocessasdiscussedabove,thewholeexponentialbackoffprocedureisalsoaMarkovprocess. LetWbetheminimumvalueofcontentionwindowsizeCWminplus1.FollowingasimilarprocedureusedbyBianchi[ 15 ]andnoticingthatthetransitionprobabilityateachbranchoftheMarkovchainisdifferentfromthere(whichonlydenotedthevalueatthesaturatedstatusanddidnotconsiderthatthecontentionwindowisresetafterthemaximumtimesofretransmissionsasdenedintheprotocols[ 68 ],wecanobtain(pleaserefertoSection 2.3.8 ) 1p+1c+(1pc)W(Pi=0(2pc)i)2(1p+1c) 1p+1c+pcWPm1i=0(2pc)i+W(12mp+1c);6m;>m9>>>>=>>>>;(2.18) wheremisthemaximumnumberofthestagesallowedintheexponentialbackoffpro-cedure(thedenitionisclariedbelow).WewilluseEquations( 2.15 )and( 2.18 )inthequeueinganalysistoderivethecollisionprobabilityatdifferentinputtrafcinSection 2.4 2 .InFig. 2 ,fs(t),b(t)gisthestateofthebi-dimensionaldiscrete-timeMarkovchain,whereb(t)isthestochasticprocessrepresentingthebackofftimercountforagivenstation,ands(t)isthestochasticprocessrepresentingthebackoffstagewithvalues(0,...,)forthestationattimet.Letmbethemaximumbackoffstageatwhichthecontentionwindowsizetakesthemaximumvalue,i.e.,CWmax=2m(CWmin+1)-1.Atdifferentbackoffstagei2[0,],thecontentionwindowsize

PAGE 38

CWi Figure2: Generalizedstatetransitiondiagramfortransmissionprocess Aswedenedbefore,therandomvariableTSisthedurationoftimetakenforastatetransitionfromthestartstate(beginningtobeserved)totheendstate(beingtransmittedsuccessfullyordiscardedaftermaximumtimesretransmissionfailures).Thus,itsProb-abilityGeneratingFunction(PGF),denotedasB(Z)thatisthefunctionofpc,nandZ,is 68 ]

PAGE 39

simplythesignaltransferfunctionfromthestartstatetotheendstategivenby: SinceB(Z)canbeexpandedinpowerseries,i.e., wecanobtainthearbitrarynthmomentofTSbydifferentiation(hencethemeanvalueandthevariance),wheretheunitofTSisslot.Forexample,themeanisgivenby whereistheMAClayerservicerate. 2 showstheprobabilitydis-tributionoftheMACservicetimeateachdiscretevalue.ThisexampleusesRTS/CTSmechanisms.ThelengthsofRTS/CTS/ACKconformtoIEEE802.11MACprotocol.Datapacketlengthis1000bytesanddatatransmissionrateis2Mbps.Thevaluesofthepara-metersaresummarizedinTableI. Wenoticethattheenvelopeoftheprobabilitydistributionissimilartoanexponentialdistribution.Ifweusesomecontinuousdistributiontoapproximatethediscreteone,itwillgiveusgreatconveniencetoanalyzethequeueingcharacteristics.Fig. 2 showstheapproximateprobabilitydensitydistribution(PDF)ofTSandseveralwell-knowncontinu-ousPDFsincludingGammadistribution,log-normaldistribution,exponentialdistribution

PAGE 40

(b) (c) (d) (e) (f) ProbabilitydistributionofMAClayerservicetime

PAGE 41

Table2: IEEE802.11systemparameters ChannelBitRate 2Mbit/s PHYheader 192bits MACheader 224bits Packetpayloadsize 1000Bytes LengthofRTS 160bits+PHYheader LengthofCTS 112bits+PHYheader LengthofACK 112bits+PHYheader Initialbackoffwindowsize(W) 31 Maximumbackoffstages(m) 5 Shortretrylimit 7 Longretrylimit 4 andErlang-2distribution.Weobservethatthelog-normaldistributionprovidesagoodapproximationforalmostallcases(notonlyforcasesatthehighcollisionprobabilitybutalsoforcasesatthelowcollisionprobability),andalsohasaveryclosetaildistributionmatchwiththatofTS.Inaddition,theexponentialdistributionseemstoprovidearea-sonablygoodapproximationexceptforcasesatverylowcollisionprobability,whereitismorelikeadeterministicdistribution.Here,thePDFofTSisobtainedbyassumingthattheprobabilitydensityfunctionisuniforminaverysmallintervalandisrepresentedbyahistogramwhileothercontinuousPDFisdeterminedbythevalueofmeanand/orvarianceofTS.Here,weuse5msastheintervalinthehistogrambecausethedistributionofthedelayconcentratesaroundtheintegertimesofthesuccessfultransmissionperiodforeachpacketwhichapproximates5msforpacketswith1000byteslong. Wealsonoticethatpchasdifferentsaturationvaluesfordifferentn.Ifthemobilestationalwayshaspacketstotransmit,i.e.,inthesaturationstate,theidleprobabilityp0takestheminimumvalue0.So,accordingtoformulae( 2.15 )and( 2.18 ),wecanobtainthesaturationvalueofpcbysettingp0as0inTableII. Table2: Saturationvalueofcollisionprobability n 5 9 17 33 65 Maxpc 0.2727 0.3739 0.4730 0.5692

PAGE 42

Figure2: PDFofservicetime Fig. 2 showsthedistributionofTSatdifferentnumberofmobilestations,whichmainlydependsonpcandhardlydependsonn.Fig. 2 showsthemeanvalueofTSatdifferentcollisionprobability.ThemaximumofTSfordifferentn,whichisreachedwhenpctakesthesaturationvalue,ismarked.WeobservethatthedistributionofTSmainlydependsonpcandisdeterminedbythenumberoftheactivestationsatsaturationstatuswhenpcreachesthesaturationvalue.Wewilldiscusshowtoobtainthevalueofpcatdifferenttrafcloadinthefollowingsection. Figure2: Meanofservicetime

PAGE 43

transmit.Wefollowthesimilarnotationsinthepaper[ 15 ].fs(t),b(t)gandWihavebeendenedinsection 2.4.6 .F.LetPfi1;k1ji0;k0gbetheshortnotationofone-steptransitionprobabilityandPfi1;k1ji0;k0g=Prfs(t+1)=i1;b(t+1)=k1js(t)=i0;b(t)=k0g.Thentheonlynonnullone-steptransitionprobabilitiesare Theseequationsaccountforthefactsthat:thebackofftimerisdecremented;theback-offtimerstartsfromstage0afterasuccessfultransmission;thebackofftimerstartsfromanewstageafteranunsuccessfultransmission;thecontentionwindowsizeisresetandthebackofftimerstartsfromstage0whenthemaximumtransmissionfailurelimitisreached,respectively. Letbi;k=limt!1Prfs(t)=i;b(t)=kg;06i6;06k<>:b;0+(1pc)P10bj;0i=0pcbi1;00
PAGE 44

Thus,b0;0canbenallydeterminedbyimposingthenormalizationcondition,whichsimpliesasfollows: Wi=Pi=0bi;0Wi+1 2=b0;0 2=b0;0 26mm1Pi=0pic2iW+1 2+Pi=mpic2mWi+1 2>m(2.26) Asanytransmissionoccurswhenthebackofftimecounterequalszero,regardlessofthebackoffstage,theprobabilitythatastation,whichhaspacketstotransmit,transmitsinarandomlychosenslottimeis whichcanbesimpliedas 1p+1c+(1pc)W(Pi=0(2pc)i)2(1p+1c) 1p+1c+pcWPm1i=0(2pc)i+W(12mp+1c);6m;>m9>>>>=>>>>;(2.28) 2.4.1Problemformulation Aqueuemodelcanbecharacterizedbythearrivalprocessandtheservicetimedistri-butionwithcertainservicediscipline.WehavecharacterizedtheMAClayerservicetimedistributionintheprevioussection.Inthischapter,weassumethatthepacketarrivalsateachmobilestationfollowthePoissonprocessoradeterministicdistributionwithaveragearrivalrate.Thepackettransmissionprocessateachstationcanbemodeledasageneral

PAGE 45

singleserver.ThebuffersizeateachstationisK.Thus,thequeueingmodelforeachstationcanbemodeledasanM/G/1/KwhenPoissonarrivalsofpacketsareassumed. whereXndenotesthenumberofpacketsseenuponthenthdeparture. Toobtainpij,wedene Moreover,wenoticethat whereB(e)isobtainedbyreplacingZwitheinequation( 2.19 ),i.e.,thePGFoftheMAClayerservicetimeTs. Accordingtothebalanceequation:

PAGE 46

where=fngandthenormalizationequation,wecancomputethe.ForthenitesystemsizeKwithPoissoninput,wehave[ 53 ] whereisthetrafcintensityand=E[TS]. IfwecanapproximatethedistributionofMACservicetimebyanexponentialdistri-bution,thesteady-stateprobabilityfortheM/M/1/Kmodel[ 53 ]isgivenby: 2.15 )and( 2.18 )usingsomerecursivealgorithm.Thus,wecanobtainthedistributionofMACservicetimeatdifferentofferedloadaccordingtotheresultsobtainedinsection 2.4.6 .Hereweassumethatthepacketarrivalprocessateachstationisindependentandidenticaldistributed,andhencewecouldobtaintheaggregateperformanceofwirelessLANfromthequeueinganalysisinthissection. (1PB)(2.35)

PAGE 47

wherep+1cisthepacketdiscardprobabilityduetotransmissionfailures. 2 showstheresultsforthemajorperformancemetrics.Allofthemhaveadramaticchangearoundthetrafcloadof1.1-1.5Mbits/sec.Thisisbecausethatthecollisionsincreasesignicantlyaroundthistrafcload,resultinginmuchlongerMACservicetimeforeachpacket. Figure2: Queuecharacteristics Fromtheresults,weobservethatallthemetricsaredependentonthecollisionprob-abilitypc.Fig. 2 showsthatpcmainlydependsonthetotaltrafcinthenon-saturatedscenario.Ontheotherhand,pcisaffectedbythetotalnumberofpacketsattemptingtotransmitbyallneighboringstations.Inthenon-saturatedcase,whenallarrivingpacketsareimmediatelyservedbytheMAClayer,thequeuelengthmayreachzeroandthecor-respondingstationwillnotcompeteforthemedium.However,inthesaturatedscenario,i.e.,thestationsalwayshavepacketstotransmit,thetotalnumberofpacketsattemptingtotransmitequalstothetotalnumberofneighboringstations,hencepcismainlydependentonthetotalnumberofneighboringstationsasweexpect.

PAGE 48

TheMAClayerservicetimeshowssimilarchangeatdifferentofferedload,becauseitisdependentonthepc.AllotherperformancemetricsaredependentonthedistributionoftheMAClayerservicetime,sotheyshowthesimilarchangeinthegures.Theaveragequeuelengthisalmostzeroatthenon-saturatedstateandreachesalmostmaximumlengthatthesaturatedstate.Theaveragewaitingtimeforeachpacketinthequeuealmostequalstozeroatthenon-saturatedstateandreachesseveralsecondsatthesaturatedstate.Thequeueblockingprobabilityiszeroatthenon-saturatedstatewhenthetrafcloadislow,andlinearlyincreaseswiththeofferedloadatthesaturatedstate.Thethroughputlinearlyincreaseswiththeofferedloadatthenon-saturatedstateandmaintainsaconstantvaluewithdifferenttotalnumberoftransmittingstationsatthesaturatedstate.Thethroughputatsaturatedstatedecreaseswhenthenumberofstationsincreasesbecausecollisionprobabil-ityclimbsupwiththenumberofstations.ThisisconsistentwiththeresultsofsaturationthroughputfoundbyBianchi[ 15 ]wheretheauthorindicatesthatthesaturatedthroughputdecreasesasnincreasesunderasmallinitialsizeofthebackoffwindowgivenaspecicsetofsystemparameters.Inaddition,thepacketdiscardingprobabilityatMAClayerismuchsmallerthanthequeueblockingprobability. Insummary,alltheseresultsindicatethatIEEE802.11MACworkswellinthenon-saturatedstateatlowtrafcloadwhileitsperformancedramaticallydegradesatthesat-uratedstate,especiallyforthedelaymetric.Besides,atthenon-saturatedstate,theper-formanceisdependentonthetotaltrafcandindifferenttothenumberoftransmittingstations.Atthesaturatedstate,thenumberoftransmittingstationsismuchmoreimportanttothewholeperformance.ThesimilarphenomenahavebeenobservedforthedistributionofMACservicetimeshowninsection. 2.5.1SimulationEnvironments 41 ].Thewirelesschannelcapacityissetto2Mbps.Datapacketlengthis1000bytes,andthemaximumqueuelengthis50.The

PAGE 49

radiopropagationmodelisTwo-RayGroundmodel.Weusedifferentnumbersofmobilestationsinarectangulargridwithdimension150mx150mtosimulatetheWirelessLAN.Allstationshavethesamerateofpacketinputs.TheMACprotocolusestheRTS/CTSbased802.11MACandotherparametersaresummarizedinTableI. 2 showsthesimulationresultsoftheMAClayerservicetimeinthenetworkwith17mobilestationsandtotaltrafcof0.2,0.8and1.6Mbps,respectively.Itdisplaysgoodmatchontheprobabilitydensityfunctionsbetweentheanalyticalresultandthatfromsimulation.Noticethat,similarlywithFig. 2 ,thePDFsshowninFig. 2 arehistogramapproximationsofthediscreteprobabilitydistributionobtainedfrombothanalysisandsimulations. Figure2: MAClayerpacketservicetime OurresultsindicatethedistributionofMAClayerservicetimeisindependentofthepacketinputdistributionwhetheritisdeterministicorPoissondistributed.Itmainlyde-pendsonthetotaltrafcinthenetworkbeforesaturationandonthenumberofmobilestationsaftersaturation,whichisconsistentwiththeanalysis.

PAGE 50

performance,suchasthroughput,linkdelay,packetdiscardingratio.Theproblemishowgoodthisapproximationisforourmodeling. Aswesaidinsection 2.4 ,theexponentialdistributionseemstobeagoodapproxima-tionfortheMAClayerservicetime.InFig. 2 and 2 ,wecompareitwiththederiveddiscreteprobabilitydistributioninthequeueinganalysistocheckitsgoodnesstopredicttheMACthroughput,packetwaitingtime,queueblockingprobabilityandaveragequeuelength.Here,weassumethatthesystemhasPoissonarrivals.Weusetwoqueueingmodelsforthesetwodistributions:M/M/1/KandM/G/1/K.Fig. 2 and 2 showtheresultsfortheWLANwith9mobilestations. (b) (c) (d) ComparisonsbetweenM/G/1/K,M/M/1/Kmodelsandsimulation

PAGE 51

Figure2: Averagewaitingtimeinnon-saturatedstatus FromFig. 2 and 2 ,weobservethatM/M/1/KmodelgiveacloseapproximationtotheM/G/1/Kmodelforsomeperformancemetrics.Bothmodelshavealmostthesamethroughputandqueueblockingprobability.However,whenthemobilestationsareatthesaturatedstate,M/M/1/Kgivesalargepredictionerrorfortheaveragequeuelengthandaveragewaitingtime,andthedifferenceissmallexceptattheturningpointbetweennon-saturatedstateandthesaturatedstate,whereadramaticchangeofthesystemperformanceisshown.TheM/G/1/Kmodelalwaysprovidesbetterapproximationforallperformancemetrics. WealsocomparetheresultsofqueueingmodelswiththesimulationinFig. 2 and 2 .Twoqueueingmodelsshowverycloseapproximationswiththesimulationresultsforallperformancemetricswhenmobilestationsareinthenon-saturatedstate.However,therearedistinctdifferencesbetweenthemwhenthesystemisinthesaturationstate.ThisisbecausethattheMarkovchainmodeloverestimatestheaverageMAClayerservicetimeabout10%inthesaturationstatecomparedtothesimulationresultsfromns-2,asshowedinFig. 2 .ThereasonsmaybethattheMarkovchainmodeldoesnotcapturealltheprotocoldetailsand/ortheimplementationconsiderationsofIEEE802.11MACprotocolsinns-2.Thus,thesimulationresultshavehigherthroughput,lowerqueueblockingprobability,smalleraveragequeuelengthandsmalleraveragewaitingtimeatsaturatedstate.

PAGE 52

Figure2: AverageMAClayerservicetime Withextensivesimulationsfordifferentnumberofmobilestationsinrandomlygen-eratedwirelessLANs,wehaveconcludedthattheMarkovchainmodelsseemtoalwaysgiveanupperboundoftheaverageMAClayerservicetime.Thus,thequeueingmodelsusingthedistributionoftheservicetimegivealowerboundofthethroughput,andupperboundsofqueueingblockingprobability,averagequeuelengthandaveragewaitingtimecomparedwithsimulationsofns-2.Therefore,ouranalyticalmodelscanalwaysbeusefultocomeupwiththeperformanceestimatesfordesignpurpose.

PAGE 53

tothenumberoftransmittingstations,andatsaturatedstate,thenumberoftransmittingstationsaffectstheperformancemoresignicantly. Inaddition,theanalyticalresultsindicatethatexponentialdistributionmayprovideagoodapproximationfortheMAClayerservicetimeinthequeueinganalysis.ThequeueingmodelsdiscussedinthischaptercanaccuratelyestimatevariousperformancemetricsofWLANinthenon-saturatedstatewhichisthedesiredstateforsomeapplicationwithacertainQoSrequirementbecausethereisnoexcessivequeueingdelayasthatinsaturatedstate.AndforWLANsinthesaturatedstate,thequeueingmodelsgivealowerboundforthethroughput,andupperboundsforqueueingblockingprobability,averagequeuelengthandaveragewaitingtimecomparedwithsimulationresultsobtainedfromns-2.

PAGE 54

ThischapterstudiesanimportantproblemintheIEEE802.11DCFbasedwirelessLAN:howwellcanthenetworksupportqualityofservice(QoS).Specically,weanalyzethenetwork'sperformanceintermsofmaximumprotocolcapacityorthroughput,delay,andpacketlossrate.Althoughtheperformanceofthe802.11protocol,suchasthrough-putordelay,hasbeenextensivelystudiedinthesaturatedcase,wedemonstratethatthemaximumprotocolcapacitycanonlybeachievedinthenon-saturatedcase,andisalmostindependentofthenumberofactivenodes.Byanalyzingthepacketdelay,consistingoftheMACservicetimeandwaitingtime,wederiveaccurateestimatesfordelayanddelayvariationwhenthethroughputincreasesfromzerotothemaximumvalue.Packetlossrateisalsogivenforthenon-saturatedcase.Furthermore,weshowthatthechannelbusynessratioprovidespreciseandrobustinformationaboutthecurrentnetworkstatus,whichcanbeutilizedtofacilitateQoSprovisioning.Wehaveconductedacomprehensivesimulationstudytoverifyouranalyticalresultsandtotunethe802.11toworkattheoptimalpointwiththemaximumthroughputandlowdelayandpacketlossrate.Thesimulationresultsshowthatbycontrollingthetotaltrafcrate,theoriginal802.11protocolcansupportstrictQoSrequirements,suchasthoserequiredbyvoiceoverIPorstreamingvideo,andatthesametime,achieveahighchannelutilization. 68 ]hasbeenwidelyusedinrecentyears.Itcontainstwoaccessmethods,i.e.,DistributedCoordinationFunction(DCF)andPointCoordinationFunction(PCF),withtheformerbeingspeciedasthefundamentalaccessmethod.Despiteitspopularuse,currentlyonly 35

PAGE 55

Table3: QoSrequirementsformultimediaservices Application One-waytransmissiondelay Delayvariation Packetlossrate Real-time VoIP,videoconferencing Streaming Streamingaudioandvideo upto10s E-mail,letransfer,webbrowsing minutesorhours N/A Zero *Playoutbuffer(orjitterbuffer)canbeusedtocompensatefordelayvariation 3.2 describesthe802.11protocolinmoredetail. FortheIEEE802.11wirelessLANtocontinuetothriveandevolveasaviablewirelessaccesstotheInternet,qualityofservice(QoS)provisioningformultimediaservicesiscrucial.AsshowninTable 3 ,forreal-time,streaming,andnon-real-time(orbesteffort)trafc,themajorQoSmetricsincludebandwidth,delay,delayjitter,andpacketlossrate[ 73 74 ].GuaranteeingQoSformultimediatrafc,however,isnotaneasytaskgiventhatthe802.11DCFisinnaturecontention-basedanddistributed,andhencerenderseffectiveandefcientcontrolverydifcult.Inaddition,otherproblemssuchashiddenterminalsorchannelfadingmakethingsworse.Toaddressthesechallenges,currentresearchworks([ 1 125 161 153 ]andreferencestherein)andtheenhancedDCF(EDCF)denedintheIEEE802.11edraft[ 72 31 ]tendtoprovidedifferentiatedserviceratherthanstringentQoSassurance. However,wehavenotyetwellunderstoodthequestionofhowwelltheIEEE802.11WLANcansupportQoSwhenmanyresearchersstarttobelievethatservicedifferentiationisthebestthatthe802.11canachieve.Inthischapter,weendeavortoaddressthisproblemthroughboththeoreticalanalysis(Section 3.3 )andsimulations(Section 3.4 ). Wedevelopananalyticalmodeltoassessthecapabilityofthe802.11forsupportingmajorQoSmetrics,i.e.,throughput,delayanddelayvariation,andpacketlossrate.Whilecurrentliteratureonperformanceanalysisisfocusedonthederivationofthroughputordelayinthesaturatedcase,wendthattheoptimaloperatingpointforthe802.11towork

PAGE 56

atliesinthenon-saturatedcase. InSection 3.5 ,weshowthatouranalyticalresultsarestillvalidevenwhentheeffectofchannelfadingistakenintoaccount.Also,wediscussthepossibleimplicationsarisingduetotheemploymentofaprioritized802.11DCF.Finally,Section 3.6 concludesthischapter. 3.2.1OperationsoftheIEEE802.11 11 ].

PAGE 57

packetfailsforamaximumnumberoftimes,thepacketisdropped.Toavoidcollisionsoflongpackets,theshortRTS/CTS(requesttosend/cleartosend)framescanbeemployed. NotethattheIEEE802.11MACalsoincorporatesanoptionalaccessmethodcalledPCF(PointCoordinationFunction),whichisonlyusableininfrastructurenetworkcong-urationsandisnotsupportedinmostcurrentwirelesscards.Inaddition,itmayresultinpoorperformanceasshowninpreviousresearch[ 126 145 ].Inthischapter,wethusfocusonthe802.11DCF. 15 21 44 58 66 134 154 160 ].Bianchi[ 15 ]proposedaMarkovchainmodelforthebinaryexponentialbackoffprocedure.Byassumingthecollisionprobabilityofeachnode'stransmissionisconstantandindependentofthenumberofretransmissions,hederivedthesaturatedthroughputfortheIEEE802.11DCF.Basedonthesaturatedthrough-putderivedinBianchi'smodel,FohandZuckerman[ 44 ]usedaMarkovianstatedependentsingleserverqueuetoanalyzethethroughputandmeanpacketdelay.Calietal.[ 21 ]stud-iedthe802.11protocolcapacitybyusingap-persistentbackoffstrategytoapproximatetheoriginalbackoffintheprotocol.Again,thefocusisonthesaturatedthroughput.Inad-ditiontocollisions,Hadzi-VelkovandSpasenovskitooktheeffectofframeerrorrateintoaccountintheiranalysisofsaturatedthroughputanddelay[ 58 ].Wederivedanapproxi-mateprobabilitydistributionoftheservicetime,andbasedonthedistribution,analyzedthethroughputandaveragedelay[ 154 160 ].Asnoticed,mostworkswerefocusedontheanalysisofthroughputanddelayinthesaturatedcase.Moreover,noneofthesesystemati-callyconsideredthedelayanddelayvariationinthenon-saturatedcase,letaloneobtainedaccurateestimatesforthem.

PAGE 58

13 20 85 90 ]orprovideprioritizedservice,namely,servicedifferentiation[ 1 81 114 119 125 137 ]. Basedonthework[ 21 ],Calietal.attemptedtoapproachtheprotocolcapacitybyreplacingtheexponentialbackoffmechanismwithanadaptiveone[ 20 ].KimandHoudevelopedamodel-basedframeschedulingalgorithmtoimprovetheprotocolcapacityofthe802.11[ 85 ].TwofastcollisionresolutionschemeswereproposedbyBharghavan[ 13 ]andKwonetal.[ 90 ],respectively.Theideaistousetwochannelsortoadjustbackoffalgorithmstoavoidcollisions,therebyprovidinghigherchannelutilization. Toprovideservicedifferentiation,AdaandCastelluccia[ 1 ]proposedtoscalethecon-tentionwindowandusedifferentinterframespacingormaximumframelengthforser-vicesofdifferentpriorities.Twomechanisms[ 125 ],i.e.,virtualMACandvirtualsource,wereproposedtoenableeachnodetoprovidedifferentiatedservicesforvoice,video,anddata.Bysplittingthetransmissionperiodintoareal-timeoneandanon-real-timeone,thereal-timetrafcissupportedwithQoSguarantee[ 114 ].However,theDCFmodewasdramaticallychanged.TheBlackbust[ 119 ]providedhighpriorityforthereal-timetraf-c.Unfortunately,itimposesspecialrequirementsonhighprioritytrafcandisnotfullycompatiblewiththeexisting802.11standard.Insummary,ifthesemanticsofthe802.11DCFismaintained,alltheworksmentionedabovecanonlysupportservicedifferentiation. Ourstudiescanbeconsideredtobeaconvergencebetweenthesetwothreadsofre-search;however,itimprovesonbothsides.WethoroughlystudytheQoSperformanceofthe802.11intermsofthroughput,delayanddelayvariation,andpacketlossrate.Moreover,wediscovertheoptimaloperatingpointatwhich,inadditiontoachievingthetheoreticalmaximumthroughput,the802.11WLANiscapableofsupportingstrictQoSrequirementsforthereal-timetrafc,ratherthanonlyprovidingprioritizedservice.

PAGE 59

15 68 ].Obviously,weobtainthefollowingequations: wherepiistheprobabilitythattheobservedbackofftimeslotisidle,psistheprobabilitythatthereisonesuccessfultransmission,andpcisthecollisionprobabilitythatthereareatleasttwoconcurrenttransmissionsatthesamebackofftimeslot.IfwedeneTsucastheaveragetimeperiodassociatedwithonesuccessfultransmission,andTcolastheaveragetimeperiodassociatedwithcollisions,weknow[ 68 ] forthecasewheretheRTS/CTSmechanismisused,and timeout+difs;(3.3)

PAGE 60

Figure3: Channelbusynessratioandutilization forthecasewherethereisnoRTS/CTSmechanism,where 15 ]forderivationof pi+psTsuc+pcTcolRb=1RiRs=psTsuc whereisthelengthofanemptybackofftimeslot,Riisthechannelidlenessratio,Rbisthechannelbusynessratio,andRsisthechannelutilization.OnceweobtainRs,thenormalizedthroughputsisexpressedas andtheabsolutethroughputisstimesthebitratefordatapackets. Inmostcases,wearemoreinterestedinthepacketcollisionprobabilitypobservedateachindividualnode,sinceitcanbeusedtocalculateQoSmetricsforthetrafctraversingthenode.Inotherwords,pistheprobabilitythatonenodeencounterscollisionswhenittransmits.Also,pistheprobabilitythatthereisatleastonetransmissionamongthe

PAGE 61

Table3: IEEE802.11systemparameters 2Mbps BitrateforRTS/CTS/ACK 1Mbps PLCPDatarate 1Mbps BackoffSlotTime 20s 10s 50s 192bits MACheader 224bits DATApacket 8000bits+Phyheader +MACheader RTS 160bits+Phyheader CTS,ACK 112bits+Phyheader Itcanbeseenthatthecollisionprobabilityincreaseswiththeincreaseinthenumberofneighboringnodesorinthetrafcateachofthesenodes.Inthissense,preectstheinformationaboutboththenumberofneighboringnodesandthetrafcdistributionatthesenodes. Accordingtotheaboveequations,wecanexpressRb,Rs,andsasafunctionofp,whichareshowninFig. 3 .AlltheparametersinvolvedareindicatedinTable 3 andmostarethedefaultvaluesintheIEEE802.11.InFig. 3 ,threecases,i.e.,n=5,10,and300,areconsidered.Itisimportanttonotethatforeachspecicn,thereexistsamaximumvalueofp,denotedbyMAX(p),atwhichthenetworkoperatesinthesaturatedstatus,i.e.,eachofthennodesalwayshaspacketsinthequeueandthuskeepscontendingforthechannel.Basedontheworks[ 15 154 160 ],weknowthatinsaturatedstatus,thelargerthenumberofnodes,thegreaterthecollisionprobability.Moreprecisely,MAX(p)=0:105,0:178,0:290,0:546,0:701,0:848forn=3,5,10,50,128,300,respectively.NextwepresentsomeimportantobservationsfromFig. 3

PAGE 62

Thisisnothardtounderstand.Whenthecollisionprobabilitypisverysmall,thechan-nelresourcewastedincollisionsissominorthatitcanbeignored.Third,thenormalizedthroughputalmoststaysunchangedwhenpincreasesfrom0:1to0:2,althoughitreachesthemaximumvaluearoundp=0:2.Finally,themaximumthroughputisalmostinsensitivetothenumberofactivenodes.GiventheseobservationsandthefactthatthethroughputisproportionaltoRs,wethereforecouldusethemeasuredchannelbusynessratioRbtoaccuratelyestimatethethroughputfromzerotothemaximumvalue.Notethatthisisverysimpleandusefultoeachnode:itcanmonitorthethroughputofthewholeWLANbysim-plymeasuringthechannelbusynessratio,whichcanbeeasilydonesincetheIEEE802.11isaCSMA-basedMACprotocol,workingonthephysicalandvirtualcarriersensingmech-anisms.Ontheotherhand,whenRbexceedsacertainthresholdthb,severecollisionscanbeobservedintheWLAN.Maximumthroughput 3 alsoshowsthatthethroughputbeginstodecreasewhenpisgreaterthanacertainvalue,andcoulddecreasetozerowhenpbecomesverylarge.Toensurethatthenetworkisalwaysworkingwithahighthroughput,itisimportantforustondthecriticalturningpoint,i.e.,whentheIEEE802.11willachievethemaximumthroughput,andhowthemaximumthroughputdependsonnetworkcharacteristicssuchasthenumberofnodenandtrafc.

PAGE 63

CombiningEquation( 3.1 )( 3.4 )( 3.6 ),wecanwriteRsasafunctionofp.Toobtainthemaximumthroughput,wetakethederivativeofRswithrespecttopandletitequal0: dpRs=0;(3.8) Meanwhile,weknowthatpisupperboundedbyMAX(p).Therefore,ifprootistherootofEquation( 3.8 ),weobtainthevalueofp,denotedbyp,withwhichthemaximumthroughputisachieved: ByapplyingptoEquation( 3.5 ),wegetthemaximumnormalizedthroughputoftheIEEE802.11atdifferentn,asshowninFig. 3(a) and 3(b) .Heretwoimportantpointsarenoted. 3(a) .p=0:196meansthereare5or6nodessimultaneouslycontendingforthechannel,whichcanbederivedfromtheinversefunctionofMAX(p)asshownearlier.Inaddition,themaximumthroughputachievedisnotsensitivetothenumberofnodes,n.Itisratherstableasnincreases. 3(b) ,ratherthanlettingp=pforeachn,ifwesimplyletp60:1orp60:05,theachievednormal-izedthroughputonlydropsby0:96%and4:2%,respectively,comparedtothemaximumnormalizedthroughput.Thisisaveryniceandimportantfeatureinthesensethataslongaseachnodeinthenetworkcankeepthecollisionprobabilitypbelowacertainvalue,say0.1,insteadofp,whichisdependentonn,themaximumthroughputiswellapproached.Thus,bymaintainingasmallcollisionprobabilityintheWirelessLAN,whichcanbedone

PAGE 64

(b)Maximumnormalizedthroughputwithdifferentcon-straintsoncollisionprobabilityp CollisionprobabilityandmaximumnormalizedthroughputwithRTS/CTSandpayloadsizeof8000bits throughcontrollingthetotalinputtrafcrate,wecanachievehighthroughput.ThisinfactisconsistentwithourobservationinFig. 3 ,whereRbRswhenp60:1. Notethatinadditiontoachievinghighthroughput,keepingasmallcollisionproba-bilityhelpsreducedelay.Sincethetimewastedduetocollisioncouldbeneglected,thecontentiondelayisverysmall,whichiscrucialinprovidinglowdelayforthereal-timetrafcandwillbediscussedindetailinsection 3.3.2 .Availablebandwidth

PAGE 65

Althoughitisnoteasyforeachindividualnodetoknowthecurrenttotalthroughputifitdoesnotdecodeeverythingreceived,thenodecanbeawareoftheavailablebandwidthbyvirtueofthechannelbusynessratio,whichcouldbeeasilyacquiredasdescribedearlier.Especially,whenp60:1,RbRs.ThusBWacanbecalculatedasfollows: whereBWisthetransmissionrateinbits/sfortheDATApackets,andthbisathresholdofRbandproportionaltothemaximumthroughput.ImpactofpayloadsizeandtheRTS/CTSmechanism 3 presentstheanalyticalresults,whereRTS/CTSisusedornotused,andvariouspayloadsizesareconsidered. WendnomatterwhetherRTS/CTSisused,thethroughputincreasesalongwiththepayloadsize.Butthisisnotnecessarilytrueforchannelutilization,suchasthecasethatRTS/CTSisnotusedinthesaturatedcase.Thereasonisthefollowing.Inthesat-uratedcase,givenn,pisxed.AccordingtoEquation( 3.1 )( 3.3 )( 3.4 )( 3.6 ),RsisalmostunchangedandRsps ItalsocanbeobservedthatthemaximumthroughputishigherinthecasethatRTS/CTSisnotusedthaninthecasethatRTS/CTSisused,nomatterhowlargethepayloadsizeis.Thisisbecausethatthemaximumthroughputisobtainedwhenpisrelativelysmallandthustheimpactofcollisionsduetolongdatapacketscouldbeignored.Asaresult,ifRTS/CTSisnotused,theMACoverheadisreduced,whichresultsinhigherthroughput.Onthecontrary,inthesaturatedcasewherethecollisionprobabilityismuchhigher,theuseofRTS/CTSdoesimprovethethroughput,especiallywhenthepayloadsizeislarge.Thisisbecausetheimpactofcollisionsduetolongdatapacketsbecomessignicantin

PAGE 66

Figure3: ImpactofpayloadsizeandtheRTS/CTSmechanism thesaturatedcaseandcannotbeignored;theexchangeofRTS/CTSavoidslongpacketcollisionsandthusreducesMACoverhead.Notethatforthepayloadsizethatisshorterthanabout220bytesinthisparametersetting,theuseofRTS/CTSiscounterproductivebecauseofitsrelativelyhighoverheadcomparedwiththeshortpayloadsize. Tosumup,tomaximizethesystemthroughput,thebasicaccesswithouttheRTS/CTSmechanismisdesired,aslongaswecankeepthecollisionprobabilityatarelativelysmallvalue.

PAGE 67

Inthefollowing,wewillgiveananalysisoftheservicetimeandthequeueingdelay.Then,theestimatesofdelayanddelayvariationarederived.Servicetimedistribution MarkovChainModelfortheServiceTime 3.2.1 ,wecancon-cludethattheonlyoutsidefactoristhecollisionprobabilitypwhenthenodeattemptsthetransmission.Asdiscussedintheprevioussection,pisdeterminedbythenumberofneighboringnodesandthetrafcdistributionatthosenodes.Thuswecouldassumethatpisindependentofthebackoffstateofthenodeunderconsideration,althoughitisstilldependentonthebackoffstatesofothernodes.WethereforecanmodelthestochasticprocessoftheservicetimeasaMarkovchain,sincethefuturestateonlydependsonthecurrentstate.Clearly,thetransitionprobabilitiesaredependentonthecollisionprobabilityp,thustheservicetimedistributionisafunctionofp. ByapplyingthesignaltransferfunctiontothegeneralizedstatetransferdiagramofMarkovChain,wehavederivedthePGFoftheservicetime,GTs(Z),whichisquiteaccu-rateasveriedbyns-2simulations([ 154 160 ]).Ontheotherhand, wheretsi(i>0)areallpossiblediscretevaluesofservicetimeTsandpi=PrfTs=tsig.Wealsofoundthatgivenp,theservicetimedistributionisalmostinsensitiveto

PAGE 68

Figure3: Meanandstandarddeviationofservicetime 3(a) .Thus,thefollowingdelayanalysisisvalidfordifferentnandweneednotspecifythevalueofn. 3.11 ,itiseasytoobtainanymomentoftheservicetimeTsbytakingthederivativeofGTs(Z)withrespecttoZ.Specically,themeanandvarianceare @ZGTs(Z)Z=1=G0Ts(1)VAR[Ts]=G00Ts(1)+G0Ts(1)[G0Ts(1)]2(3.12) Fig. 3 demonstratesthemeanandvarianceoftheservicetimeasafunctionofthecollisionprobabilityp.Itcanbeseenthatwhenp>0:1,boththemeanandthevarianceincreaseexponentiallywithp.Ontheotherhand,wehavefoundthatwhenp60:1,theachievedthroughputisalmostthesameasthemaximumachievablethroughput.Toprovideadelayguaranteeforsomedelay-sensitiveapplicationssuchasvoiceoverIP,andachieveapproximatelymaximumthroughput,thewirelessLANshouldkeepthecollisionprobabilitylessthan0:1.Packetdelayboundanddelayvariationestimate

PAGE 69

ofallapplicationsthatemitpacketstotheMAClayer;theservicetimefollowsthedis-tributiondescribedinprevioussubsection.Afterbuildingsuchaqueueingmodel,wecanderiveaccurateestimatesofdelayanddelayvariationinthenon-saturatedcase.Noticethatthenumberofpacketswaitinginthequeue,Nq,almostequalszerointhenon-saturatedcaseespeciallyforp60:1asshowninthepapers[ 154 160 ]andveriedinoursimulationlater.Otherwiseeachnodewillcontendforthechannelinmosttimesandresultinamuchhigherp. 86 ].Accordingly,themeanofthepacketdelayT,whichconsistsofthewaitingtimeinthequeueandtheservicetime,is Ts2 whereistheaveragearrivalrateoftheinputtrafcand= 87 ], 2(1)TU(3.14) whereTsandaarethestandarddeviationsoftheservicetimeandpacketarrivalprocess,respectively. Sofar,theseresultsholdwhen<1forthesystemwithinnitebuffer.TheactualdelayupperboundshouldbelessthanTUbecausewedonotcountthepacketsdroppedduetolimitedbuffer,whichwillhavealongdelayinthesystemwithinnitebuffer.Infact,becauseweareonlyinterestedinthenon-saturatedcasewithanalmostemptyqueue,theaboveresultsarethusaccurate.

PAGE 70

InthepreviousparagraphweonlygivethemeanofthepacketdelayTwiththees-timationdependentonthespecicpacketarrivalprocessandontheaccurateestimateof.Inreality,however,thisapproachcouldbeinfeasibleifishardtoestimatewhentheinstantaneouspacketarrivalrateateachindividualnodechangesdramatically.Wethusembarkonderivingtheaccurateestimatesfordelayanddelayvariationinamoregeneralcase,i.e.,withoutanyknowledgeabout. LetTsidenotetheservicetimeforthei-thpacketatanodeunderconsideration.Sincethebackofftimerisresetforeverypackettobetransmitted[ 68 ],fTs1;Ts2;:::gareiid(independentlyandidenticallydistributed)randomvariables.LetTibethesystemtime(ordelay)ofthei-thpacketincludingtheservicetimeandthewaitingtimeinthequeue,Ribetheresidualservicetimeseenbythei-thpacket,andNibethenumberofpacketsfoundwaitinginthequeuebythei-thpacketattheinstantofarrival. Basedonsuchnotations,weobtain Aspreviouslydiscussed,Nialmostequals0inthenon-saturatedcase,sowecanapproximateTias NoticethatRiistheresidualservicetimeofthe(iNi1)-thpacket,thuswehave BytakingexpectationsonbothsidesofEquation( 3.17 ),wehave

PAGE 71

SinceitisdifculttoderivethevariationofRiingeneralcases,weusethestandarddevi-ationoftheservicetimeTstoapproximatethatofTi,i.e.,Tasfollows: wherekisaconstantvalue.Fromns-2simulationresultsaspresentedlater,k=1,or2givesagoodapproximation. Infact,byapplyingtheResidualLifeTheorem[ 86 ],wecouldobtainmoreaccurateapproximationsofE[T]andT.Letrbetheresidualservicetimeobservedatanytimeinstantduringtheservice.IftheservicetimedistributionisFTs(x),thenthepdfofr,denotedbyfr(x),canbeexpressedas(1FTs(x)),where=1 wherer0=P(busy)istheprobabilitythattheserverisbusy,i.e.,thereisonepacketcontendingforthechannelorbeingtransmitted.Becauser061,weobtain 2E[Ts]TUR(3.21) 12E[Ts](E[Ts2] 2E[Ts])22TUR(3.23)

PAGE 72

Fig. 3(a) illustratesboththelowerboundandtheupperboundforthepacketdelayT.Wecanseethattheupperboundandlowerboundareveryclose,thuswecancharac-terizethedelaywithhighaccuracy,althoughtheexactvalueisnotavailable.Asexpected,whenp<0:1,TURistighterthan2E[Ts].Thisisdesirablesincewefocusonthenon-saturatedcasewherepissmall.Asrevealedbythebounds,themeanofthesystemdelayTissmall:5ms0:5;however,r0shouldhavebeensmallerthan0.5whenp60:002. Asaspecialcase,ifthepacketarrivalprocessisPoissonian,thenr0==E[Ts]<1.Thus 2E[Ts2]TURM;(3.24) 3E[Ts3](1 2E[Ts2])22TURM:(3.25) Finally,wecommentontheresultsofdelayanddelayvariation.First,alltheaboveresultsarederivedforthenon-saturatedcase,whichmeansthetrafcintensity<1andthecollisionprobabilityp60:1.Second,theapproximationinEquation( 3.16 )reliesontheassumptionthatthereisnobulkarrival.AlthoughthisassumptioniscommonintheanalysisofqueueingsystemsandistrueforboththePoissonarrivalprocessanddetermin-isticarrivalprocess,inpractice,burstytrafcsuchasTCPtrafcviolatesit.Consequently,theburstytrafcinducesnotonlylongerwaitingtimeinthequeue,butalsohighercolli-sionprobabilityintheburstperiodleadingtolongerservicetime.Fortheaboveresultstoremainvalid,itisnecessarytoregulatearrivingtrafcattheMAClayer.

PAGE 73

(b)Standarddeviationofdelay Packetdelay WhenthepacketblockingprobabilityPblock,i.e.,theprobabilitythatthequeueisfullwhenapacketarrives,isverysmall,asforthenon-saturatedcasewhereNq=0asmentionedearlier,thetotalpacketlossratePlofthequeueingsystemcanbeapproximated

PAGE 74

asPd,i.e., Weseewhenp60:1and=7[ 68 ],Pl6107.Obviously,thissatisesthepacketlossrequirementsofmostapplicationssuchasVoIP. Onthecontrary,amuchhigherpacketlossrateisexpectedifthenetworkisinthesaturatedcaseforthefollowingreason.Ononehand,thecollisionprobabilitypgetssig-nicantlylarge,resultinginconsiderablepacketlossesduetocollisions.Ontheotherhand,eachpacketexperiencesamuchlongersystemdelayinthesaturatedcasecomparedtothatinthenon-saturatedcase,whichleadstoafullqueueatmosttimesandhenceblocksnewlyarrivingpackets. Beforeendingthissection,wemakeafewremarksabouttheanalyticalmodel.Notethatalltheperformancemetricsareexpressedasafunctionofthecollisionprobability.However,obtainingthecollisionprobabilityisnoteasy.Therearetwopossibleapproaches.Oneistoanalyticallyderivethecollisionprobability,whichrequiresthefullknowledgeofthetrafcarrivalmodelsatthenodeofinterestandatalltheothernodesinthenetworkaswell.Theotheristomeasureitthroughexperiments.Unfortunately,itisnotamenabletopracticalmeasurementduetothelackofmeasuredvaluesortheinabilityofeachnodetodistinguishcollisionsfromchannelfading.Therefore,weproposethechannelbusynessratioasagoodsubstituteforthecollisionprobabilityforthefollowingreasons.First,asmentionedearlier,thechannelbusynessratioisaninjectivefunctionofthecollisionprobability.Thisindicatesthatthechannelbusynessratiocanalsoserveastheinputoftheanalyticalmodel.Unlikethecollisionprobability,thechannelbusynessratioiseasytomeasureinpracticebecausetheIEEE802.11isessentiallybasedoncarriersensing.Second,asshownforthenon-saturatedcase,thechannelbusynessratiocanaccuratelyrepresentthechannelutilizationorthenormalizedthroughput,andhencecanbeusedtofacilitatenetworkcontrolmechanismssuchascalladmissioncontroloverthereal-time

PAGE 75

trafcandratecontroloverthebestefforttrafc.Accordingly,alltheperformancemetricsarepresentedasafunctionofthechannelbusynessratiointhefollowingsimulationresults. 3.3 .Second,whileouranalyticalresultshaveshownthattheIEEE802.11canoperateatanoptimalpointthatleadstomaximumthroughput,lowdelay,andalmostzeropacketlossrate,theydonotrevealaspecicwaytoachievethisoptimaloperatingpoint.Thuswedemonstratehowtoreachandretaintheoptimalpointthroughsimulations. 3 .TheRTS/CTSmechanismisused.WesimulatedifferentnumberofmobilestationsinthewirelessLAN.EverynodeinitiatesanidenticalUDP/CBRtrafcowtoarandomlyselectedneighbor.Thequeuelengthateachnodeis10packets. Asrevealed,whetherthenetworkoperatesinthenon-saturatedorsaturatedcasecanbedeterminedbycontrollingthecollisionprobabilityp.Also,theoptimaloperatingpointlieswherep0:1.Withoutchangingthe802.11protocol,weusetwotechniquestocontrolpinordertolocatetheoptimalpoint.OneistoschedulethestarttimeoftheUDPows,whichwillbedescribedbelow;theotheristograduallyincreasethesendingrateofeachowfrom0.Incontrast,thesaturatedcasecanbeeasilysimulatedbyboostingthetrafcloadtoamuchhigherlevelthanwhatthenetworkcansupport.Deterministicminimum-collision-probabilityscheduling(DPS)

PAGE 76

Figure3: Simulationresultswhenpayloadsize=8000bits collisionprobabilitycouldbereducedtozero.Inthiscase,thereisnoqueueingdelayandthesystemdelayistherandombackofftimeplusonepackettransmissiontime.Wecallthisschedulingdeterministicminimum-collision-probabilityscheduling.Distributedrandomizedscheduling(DRS)

PAGE 77

Figure3: Simulationresultswhenn=50andpayloadsize=8000bits 3 ,forthenon-saturatedcase,weseethatthenormalizedthroughputthatDPSachievesisslightlyhigherthanthetheoreticalmaximumthroughput,sinceitusesperfectschedulingandhencereducesthecollisionprobabilitytozero.Likewise,thenor-malizedthroughputthatDRSachievesisclosetothetheoreticalmaximumthroughput,sinceitgreatlyreducesthecollisionprobability.Onthecontrary,thethroughputinthesaturatedcaseismuchlower.Asisconsistentwiththeanalyticalresults,thenon-saturatedthroughputisalmostindependentofthenumberofnodes,whereasthesaturatedthrough-putdeclinessignicantlywiththeincreaseinnodenumber.Fordelay,weseethatthereisdifferenceinordersofmagnitudeforthesetwocases.Also,whilethedelaystaysalmostunchangedinthenon-saturatedcaseasthenumberofnodeincreases,itincreasesinthesaturatedcase.Thisisduetothefactthatinthelattercase,eachnodealwayshaspacketstotransmitandkeepscontendingforthechannel,whichgreatlyincreasesthecollisionprob-ability.Asaresult,eachpacketsuffersfrombothlongqueueingdelayandservicetime.NotethatDPSenjoysashorterdelaythanDRSsinceitreducesthecollisionprobabilitymoreeffectively.Optimaloperatingpoint 3 shows,DRSyieldsacomparableperformancewiththatofDPS,wethususeDRSasourschedulingalgorithmhenceforth.Bygraduallyincreasingthesendingrate

PAGE 78

ofeachow,weareabletolocatetheoptimaloperatingpointasshowninFig. 3 and 3 .WhileFig. 3 presentstheperformanceofthroughput,delay,anddelayvariationasafunctionofthechannelbusynessratio,Fig. 3 showsthebehaviorofaveragequeuelengthandpacketlossratewheninputtrafcincreases. Twoimportantobservationsaremade.First,weobservethereisaturningpointinallthecurveswherethechannelbusynessratioisabout0.95.Beforethatpoint,astheinputtrafcincreases,thethroughputkeepsincreasing,thedelayanddelayvariationaresmallandalmostunchanged,thequeueateachnodeisempty,andthepacketlossrateiszero.Notethatwiththesmalldelayanddelayvariation,thedelayrequirementsofthereal-timetrafccanbeadequatelysupported.Afterthatpoint,thequeueandthecollisionprobabilityformsapositivefeedbackloop.Aslightlylargercollisionprobabilitycausesthequeuetobuildup.Thequeue,evenwithonepacketalwaysinit,willforcetheMACtokeepcontendingforthechannel,therebyexponentiallyincreasingthecollisionprobability,whichinturnforcesmorepacketstoaccumulateinthequeue.Then,catastrophiceffectstakeplace:thethroughputdropsquickly,thequeuestartstobuildupandthedelayanddelayvariationincreasedramatically,andthepacketsuffersfromalargelossrate.Clearly,thisturningpointistheoptimaloperatingpointthatweshouldtunethenetworktoworkaround,wherethethroughputismaximizedandthedelayanddelayvariationaresmall. Second,asshowninFig. 3 ,thesimulationresultsverifyouranalyticalstudyoftheIEEE802.11.Thethroughputcurvesobtainedfromanalysisandsimulationcoincidewitheachother.Alsoasindicatedinouranalyticalstudy,beforetheoptimalpointisreached,thenetworkstaysinthenon-saturatedcaseandthequeueingdelayisalmostzero;thusthepacketdelayTcanbeaccuratelyestimatedbytheservicetimeTS,whichprovidesthelowerbound.Meanwhile,themeanandvariationarewellboundedbyTURandTURbeforetheturningpointasshowninequations( 3.18 3.19 3.21 3.23 ).

PAGE 79

Figure3: Simulationresultswhenn=50andpayloadsize=8000bits 3.5.1ImpactofFadingChannel 58 ]did,andalltheanalyticalresultsstillhold. ItisimportanttonotethatnormallychannelfadingisnotaseriousproblemintheWLAN,whichfeatureslownodemobilityandrelativelystablechannel.However,ifthepacketerrorprobabilityduetochannelfadingbecomessignicant,i.e.,theequivalentcol-lisionprobabilityishighinourmodel,theQoSlevelwillbehurt.Ouranalyticalresultsshowthatinthiscase,asillustratedinFig. 3 3 3(a) ,and 3(b) ,thenormalizedthroughputdecreases,theservicetimeincreases,themeanandvariationofdelayincrease

PAGE 80

alongwiththeservicetime,andpacketlossrateincreasesaswell.However,withouran-alyticalmodel,wecanstillcalculatethemaximumthroughput,packetlossrate,andgiveaccurateestimatesofdelayanddelayvariationaccordingtoEquations( 3.5 ),( 3.18 ),( 3.19 ),( 3.21 ),and( 3.23 ). 1 125 ]isadopted,weareabletoprovideprioritywithinthereal-timetrafc.Asaresult,thehighpriorityreal-timetrafcreceivessmallerdelayvariation,whereasthelowpriorityreal-timetrafcreceiveshigherdelayvariation[ 33 ]. Wehaveanalyticallycharacterizedtheoptimaloperatingpointforthe802.11WLAN,andshownthatifthenetworkistunedtoworkatthispoint,inadditiontoachievingtheoret-icalmaximumthroughput,itcansupportthemajorQoSmetricssuchasthroughput,delayanddelayvariation,andpacketlossrate,asrequiredbyreal-timeservices.Thisisfurthervalidatedviaextensivesimulations.WethereforeclarifythattheIEEE802.11WLANcanprovidestatisticalQoSguarantees,notjustdifferentiatedservice,formultimediaservices.Wealsodemonstratethatthechannelbusynessratiocanaccuratelyandtimelyrepresentthenetworkutilization;henceitcanbeusedtofacilitatetheregulationoftotalinputtrafctosupportQoS.

PAGE 81

Qualityofservice(QoS)supportformultimediaservicesintheIEEE802.11wirelessLANisanimportantissueforsuchWLANstobecomeaviablewirelessaccesstotheInternet.Inthischapter,weendeavortoproposeapracticalschemetoachievethisgoalwithoutchangingthechannelaccessmechanism.Tothisend,anovelcalladmissionandratecontrol(CARC)schemeisproposed.ThekeyideaofthisschemeistoregulatethearrivingtrafcoftheWLANsuchthatthenetworkcanworkatanoptimalpoint.Werstshowthatthechannelbusynessratioisagoodindicatorofthenetworkstatusinthesensethatitiseasytoobtainandcanaccuratelyandtimelyrepresentchannelutilization.Thenweproposetwoalgorithmsbasedonthechannelbusynessratio.Thecalladmissioncontrolalgorithmisusedtoregulatetheadmissionofreal-timeorstreamingtrafcandtheratecontrolalgorithmtocontrolthetransmissionrateofbestefforttrafc.Asaresult,thereal-timeorstreamingtrafcissupportedwithstatisticalQoSguaranteesandthebestefforttrafccanfullyutilizetheresidualchannelcapacityleftbythereal-timeandstreamingtrafc.Inaddition,theratecontrolalgorithmitselfprovidesasolutionthatcouldbeusedabovethemediaaccessmechanismtoapproachthemaximaltheoreticalchannelutilization.Acomprehensivesimulationstudyinns-2hasveriedtheperformanceofourproposedCARCscheme,showingthattheoriginal802.11DCFprotocolcanstaticallysupportstrictQoSrequirements,suchasthoserequiredbyvoiceoverIPorstreamingvideo,andatthesametime,achieveahighchannelutilization. 68 ]hasbeenincreasinglyemployedtoaccesstheInternetbecauseofitssimpledeploymentandlowcost.Accordingtothe 62

PAGE 82

IEEE802.11standard,themediumaccesscontrol(MAC)mechanismcontainstwoaccessmethods,i.e.,DistributedCoordinationFunction(DCF)andPointCoordinationFunction(PCF),withtheformerbeingspeciedasthefundamentalaccessmethod.Despiteitspopularuse,currentlyonlybestefforttrafcissupportedinDCF.Section 4.2 describesthe802.11protocolinmoredetail. Qualityofservice(QoS)provisioningformultimediaservicesincludingvoice,video,anddataiscrucialfortheIEEE802.11wirelessLANtocontinuetothriveandevolveasaviablewirelessaccesstotheInternet.Althoughthereareseveralschemes([ 96 9 88 34 124 ])whichusePCFmodetosupportQoSforreal-timetrafc,wedonotdiscussfurtheralongthislinebecausePCFisanoptionalaccessmethod([ 68 ])whichisonlyusableoninfrastructurenetworkcongurationsandnotsupportedinmostcurrentwirelesscards.Inaddition,itmayresultinpoorperformanceasshowninthepapers[ 94 145 126 ].Inthischapter,wefocusonthe802.11DCFmode.However,guaranteeingQoSforreal-timetrafcinthe802.11DCFmodeisnotaneasytaskgiventhatitisinnaturecontention-basedanddistributed,andhencerendereffectiveandefcientcontrolverydifcult.Furthermore,otherproblemssuchashiddenterminalsorchannelfadingmakethingsworse. Infaceofthesechallenges,considerableresearch([ 1 81 107 114 119 125 137 ])hasbeenconductedtoenhancetheIEEE802.11WLANtosupportservicedifferentiationorprioritizedservice[ 18 ].AdaandCastelluccia[ 1 ]proposedtoscalethecontentionwindow,usedifferentinterframespacingormaximumframelengthforservicesofdifferentpriority.Asamatteroffact,similarideashaverecentlybeenadoptedintheenhancedDCF(EDCF)denedintheIEEE802.11edraft([ 72 31 99 ]).twomechanisms[ 125 ],i.e.,virtualMACandvirtualsource,wereproposedtoenableeachnodetoprovidedifferentiatedservicesforvoice,video,anddata.Bymodifyingthe802.11MAC,adistributedpriorityschedulingschemewasdesignedtoapproximateanidealizedschedule,whichsupportsprioritizedservices[ 81 ].Similarly,bysplittingthetransmissionperiodintoareal-timeoneandanon-real-timeone,real-timetrafcissupportedwithQoSguarantee[ 114 ].However,the

PAGE 83

DCFmodewasdramaticallychanged.TheBlackbust[ 119 ]providedhighpriorityforreal-timetrafc.Unfortunately,itimposesspecialrequirementsonhighprioritytrafcandisnotfullycompatiblewiththeexisting802.11standard.Insummary,ifthesemanticsofthe802.11DCFismaintained,onlydifferentiatedservice,ratherthanstringentQoSassurance,issupported. Meanwhile,muchefforthasalsobeenspentinimprovingthroughputforthe802.11DCF([ 12 13 16 20 85 90 ]).Basedonthework[ 21 ],Calietal.attemptedtoapproachtheprotocolcapacitybyreplacingtheexponentialbackoffmechanismwithanadaptiveone[ 20 ].KimandHoudevelopedamodel-basedframeschedulingalgorithmtoimprovetheprotocolcapacityofthe802.11[ 85 ].TwofastcollisionresolutionschemeswereproposedbyBharghavan[ 13 ]andKwonetal.[ 90 ],respectively.Theideaistousetwochannelsortoadjustbackoffalgorithmstoavoidcollisions,therebyprovidinghigherchannelutilization.Itisimportanttonotethatalltheseworksfocusedonthethroughputinthesaturatedcase. Inourpreviouswork[ 150 ],WehaveshownthroughboththeoreticalandsimulationstudiesthattheIEEE802.11DCFprotocolcouldsatisfytheQoSrequirementsofthereal-timeandstreamingtrafcwhileachievingthemaximalchannelutilizationwhenitiswork-ingattheoptimalpointcorrespondingtoacertainamountofarrivingtrafc.Ifthearrivingtrafcisheavierthanthisthreshold,theWLANenterssaturation,resultinginsignicantincreaseindelayanddecreaseinthroughput;ontheotherhand,ifthearrivingtrafcislessthanthisthreshold,channelcapacityiswasted.Inreality,however,totunethenetworkthatoperatesonthebasisofchannelcontentiontoworkatthispointrequiresaneffectiveandefcientcontrolalgorithmtoregulatetheinputtrafc[ 109 ].Therefore,wearemotivatedtodesignacalladmissionandratecontrolscheme(CARC)(Section 4.4 ).Specically,calladmissioncontrol(CAC)isusedforreal-timeorstreamingtrafc,andratecontrol(RC)forbesteffortdatatrafc. Essentially,theCARCschemehasthefollowingdistinguishingfeatures:

PAGE 84

4.3 WehaveimplementedtheCARCschemeinns-2[ 106 ],andconductedacomprehen-sivesimulationstudytoevaluateitsperformance.AsshowninSection 4.5 ,CARCisabletosupportreal-timeservices,suchasvoiceandvideo,withQoSguarantees,andachievehighthroughputbyallowingbestefforttrafctomakefulluseoftheresidualchannelca-pacity.Thisconrmsthatthe802.11WLANcannotonlysupportdifferentiatedservice,butalsosupportstrictQoS. InSection 4.6 ,wediscusstheeffectofchannelfadingonourschemeandthepossibleimplicationsarisingduetotheemploymentofaprioritized802.11DCF.Finally,Section 4.7 concludesthischapter. 4.2.1OperationsoftheIEEE802.11DCFProtocol

PAGE 85

backofftimerisdecreasedbyone.Otherwise,itissuspended.Whenthebackofftimerreacheszero,thenodetransmitsaDATApacket.Ifthereceiversuccessfullyreceivesthepacket,itacknowledgesthepacketbysendinganacknowledgment(ACK)afteraninter-valcalledshortinter-framespace(SIFS).Sothisisatwo-wayDATA/ACKhandshake.Ifnoacknowledgmentisreceivedwithinaspeciedperiod,thepacketisconsideredlost;sothetransmitterwilldoublethesizeofCWandchooseanewbackofftimer,andstarttheaboveprocessagain.Whenthetransmissionofapacketfailsforamaximumnumberoftimes,thepacketisdropped.Toreducecollisionscausedbyhiddenterminals[ 14 ],theRTS/CTS(requesttosend/cleartosend)mechanismisemployed.Therefore,afour-wayRTS/CTS/DATA/ACKhandshakeisusedforapackettransmission. IntheIEEE802.11,thenetworkcanbeconguredintotwomodes,i.e.,infrastructuremodeoradhocmode.Intheinfrastructuremode,anaccesspoint(AP)isneededtopartic-ipateinthecommunicationbetweenanytwonodes,whereasintheadhocmode,allnodescandirectlycommunicatewitheachotherwithouttheparticipationofanAP. 73 74 ],theonewaytransmissiondelayshouldbeprefer-ablylessthan150ms,andmustbelessthan400ms.However,itisnotverysensitivetopacketlossrate.Typically,alossrateof1%isacceptableforreal-timevideowithrate

PAGE 86

74 ],acceptabledelaymaybeupto10seconds,whilethepacketlossrateisabout1%.StreamingtrafcisnormallytransportedviaUDP,althougharetransmissionstrategycanbeaddedintheapplicationlayer. 68 15 154 160 ]).Letpi,ps,andpcbetheprobabilitiesthattheobservedbackofftimeslotisoneofthethreekindsofslots,respectively.LetTsucbetheaveragetimeperiodassociatedwithonesuccessfultransmission,andTcolbetheaveragetimeperiodassociatedwithcollisions.Then timeout+difs=rts+eifs;(4.1)

PAGE 87

forthecasewheretheRTS/CTSmechanismisused,and timeout+difs= forthecasewherethereisnoRTS/CTSmechanism,where 15 ]forderivationof 68 ]withaneifsperiod.Thus,itcanbeeasilyobtainedthat pi+psTsuc+pcTcolRb=1RiRs=psTsuc whereisthelengthofanemptybackofftimeslot,Riisdenedasthechannelidlenessratio,Rbthechannelbusynessratio,andRsthechannelutilization.Clearly,thechannelbusynessratioRbcanalsobethoughtofastheratiooftimethatthechannelisbusyduetosuccessfultransmissionsaswellascollisionstothetotaltime.OnceweobtainRs,thenormalizedthroughputsisexpressedas andtheabsolutethroughputisstimesthebitratefordatapackets.

PAGE 88

WeassumethetotalnumberofnodesinaWLANisn.Thetransmissionprobabilityforeachnodeinanybackofftimeslotispt.Obviously,weobtainthefollowingequations: Meanwhile,pcanbeexpressedintermsofptasfollows: AccordingtoEquation( 4.3 )( 4.5 )( 4.6 ),wecanexpressRb,Rs,andsasafunctionofp,whichareshowninFig. 4 .AlltheparametersinvolvedareindicatedinTable 4 andmostarethedefaultvaluesintheIEEE802.11.InFig. 4 ,threecases,i.e.,n=5,10,and300,areconsidered. SeveralimportantobservationsaremadeforFig. 4 .First,wendthatthechannelbusynessratioisaninjectivefunctionofthecollisionprobability.Infact,thiscaneasilybeproved.Second,whenp60:1,RbisalmostthesameasRs,namely Thisisnothardtounderstand.Whenthecollisionprobabilitypisverysmall,thechannelresourcewastedincollisionsissominorthatitcanbeignored.Third,themaximalthrough-putisalmostinsensitivetothenumberofactivenodes.Asamatteroffact,wehaveshowninourpreviouswork[ 150 ]thatthepointwherethemaximalthroughputisachievedistheoptimalworkingpointforthenetworkwherethecollisionprobabilityisverysmallandthepacketdelayanddelayjitteraresmallenoughtosupporttheQoSrequirementsofreal-timetrafc.GiventheseobservationsandthefactthatthethroughputisproportionaltoRsasshowninEquation( 4.4 ),wethereforecouldusethemeasuredchannelbusynessratioRbtoaccuratelyestimatethethroughputfromzerotothemaximumvalue.

PAGE 89

Figure4: Channelbusynessratioandutilization Next,wepresentsomens-2simulationresultsinFig. 4 ,whichshowstheperfor-manceofthroughput,delay,anddelayvariationasafunctionofthechannelbusynessratio.Again,theIEEE802.11systemparametersaresummarizedinTable 4 .Everynodeiniti-atesanidenticalUDP/CBRtrafcowtoarandomlyselectedneighbor.Thequeuelengthateachnodeis100packets.DifferentpointsinFig. 4 correspondstodifferentsendingrateofows.Itcanbeseenthatthereisaturningpointinallthecurves,wherethechannelbusynessratioisabout0.95.Beforethatpoint,astheinputtrafcincreases,thethroughputkeepsincreasing,thedelay(includingqueueingdelay,backofftimeandtransmissiontime)anddelayvariationdoesnotchangemuchandissmallenoughtosupportthereal-timetrafc.Afterthatpoint,thethroughputdropsquicklyandthedelayanddelayvariationincreasedramatically.Clearly,thisturningpointistheoptimaloperatingpointthatweshouldtunethenetworktoworkaround,wherethethroughputismaximizedandthedelayanddelayvariationaresmall.Therefore,thenetworkstatusisknownbykeepingtrackofthechannelbusynessratio. Further,ifwedenotebyBUthechannelutilizationcorrespondingtotheoptimalpoint,wecanestimatetheavailablenormalizedthroughputbysa=(BURb) 150 ],BUisalmost

PAGE 90

Table4: IEEE802.11systemparameters 2Mbps BitrateforRTS/CTS/ACK 1Mbps PLCPDatarate 1Mbps BackoffSlotTime 20s 10s 50s 192bits MACheader 224bits DATApacket 8000bits+Phyheader+MACheader RTS 160bits+Phyheader CTS,ACK 112bits+Phyheader Simulationresultswhennumberofnodesequals50andRTS/CTSmechanismisused thesamefordifferentnumberofactivenodesandpacketsize,andBU0:90(withoutRTS/CTS)orBU0:95(withRTS/CTS). 68 ]indicatesthechannelisbusy,andtobeidleotherwise.

PAGE 91

overreal-timetrafcandratecontrol(RC)overbestefforttrafc,giventhatthe802.11DCFprotocolisdesignedtoprovidebesteffortservicesanddoesnotdifferentiateanytypesofservices. Wethusproposeacalladmissionandratecontrol(CARC)scheme,whichconsistsoftwomechanisms:CACandRC.Inwhatfollows,thedesignrationaleisdiscussedrst,followedbydetaileddescriptionsoftheCACandRCalgorithminorder. Therstproblemisthatmultiplenewreal-timeowsmaybesimultaneouslyadmit-tedbyindividualnodesifnotcoordinated,henceforthreferredtoasover-admission.Tomitigatethisproblem,eachnodecanrandomlybackofftodelayanewowthatcouldbeadmitted.Duringthebackoffperiod,eachnodekeepsmonitoringthechannelbusynessratio;ifthemeasuredchannelbusynessratioisincreased(duetotheadmissionofnewowsbyothernodes)suchthatthepreviouslycould-be-admittedbutdelayednewowcannolongerbeaccepted,theowisrejected.Anotherwayisthateachnode,afteradmittinganewow,dropstheowiflateronthemeasuredchannelbusynessratioisfoundtobegreaterthanthemaximumchannelutilization.Inthiscase,however,theQoSlevelofthereal-timeowsadmittedearlierhavealreadybeensuffered. Anothermoresevereissueisthatitisveryhardforeachindividualnodetoaccuratelyestimatethetotaltrafcrateofthecurrentlyadmittedreal-timeowsbasedonthemea-suredchannelbusynessratio,sincethelatteralsoincludesthecontributionfrombestefforttrafc.Withoutanaccurateestimate,therateofbestefforttrafccannotbeeffectively

PAGE 92

controlled.ThisinturnmaycompletelycausetheCACalgorithmtorejectanyreal-timetrafcifthechannelbusynessratioisboostedtoahighlevelbyheavybestefforttrafc. Therefore,toachieveitsgoal,theCARCschememustproperlyaddresstheseprob-lems.Tocompletelyavoidtheover-admissionproblem,weoptforacoordinator-aidedCACscheme.Inotherwords,alladmissiondecisionsaremadebyacoordinatingnode,whichcanrecordthecurrentnumberofadmittedreal-timeowsandtheiroccupiedchan-nelbandwidthinthenetwork.Clearly,inthiswaynoover-admissionwilloccur.Itisim-portanttonotethatacoordinatorisavailablewhetherthewirelessLANisworkingintheinfrastructuremodeorintheadhocmode.Ifthenetworkisworkingintheinfrastructuremode,theaccesspointisthecoordinator.Otherwise,amobilenodecanbeelectedtoactasthecoordinatorinthenetworkusingoneofmanyalgorithmsintheliterature([ 49 116 ]).Furtherdiscussionsontheelectionalgorithmisbeyondthescopeofthischapter. Sincethe802.11DCFisnotprioritized,ourCACalgorithmguaranteesauniformQoSlevelintermsofdelay,delayvariation,andpacketlossrateforallreal-timetrafc.NotethattwocriteriaareappliedtoCAC.TherstcriterionisthatCACadmitsanewreal-timeowonlyiftherequestedresourceisavailable.Hereweneedtosetanupperbound,denotedbyBM,forbandwidthreservationforreal-timetrafc[ 33 ].WesetBMto80%(itcouldbechangeddependingontrafccomposition)ofthemaximumchannelutilization,denotedbyBU,oftheWLANfortworeasons.Itrstensuresthatthebestefforttrafcisoperationalallthetime,sincethebestefforttrafcisatleastentitledto20%ofthechannelthroughput.Inaddition,the20%ofthechannelthroughputforbestefforttrafccanbeusedtoaccommodatesizableuctuationscausedbyVBRreal-timetrafc.ThesecondcriterionisthattheQoSprovidedforthecurrentlyexistingreal-timeowsisnotaffected.Thiscanbeguaranteedaslongastherstcriterionisinplacetomakesurethecollisionprobabilityiskeptaroundasmallvalueasshownearlier. Forbestefforttrafc,theratecontrol(RC)schememustensuretwothings.First,bestefforttrafcshouldnotaffecttheQoSleveloftheadmittedreal-timetrafc.Second,best

PAGE 93

efforttrafcshouldhaveaccesstotheresidualbandwidthleftbyreal-timetrafcinordertoefcientlyutilizethechannel.Clearly,bothdemandanaccurateestimateoftheinstan-taneousrateofongoingreal-timetrafc.Ifthenetworkisworkingintheinfrastructuremode,thisisachievable.Inthiscase,sinceallcommunicationsmustgothroughtheaccesspoint,itcanmonitorthetrafcinbothdirections,i.e.,theupstreamowsthatarefrommobilenodestotheaccesspoint,andthedownstreamowsthatarefromtheaccesspointtomobilenodes.Ontheotherhand,ifthenetworkisworkingintheadhocmode,accurateratecontrolbecomesmuchmoredifcult.Inthiscase,sinceallmobilenodescandirectlycommunicatewitheachother,nonodehasperfectknowledgeoftheinstantaneoustrafcrateofthereal-timetrafcastheaccesspointdoes.Atthesametime,nosinglenodecanaccuratelymonitorallthetrafcintheairandcontrolthetrafcrateofeveryothernode.Therefore,aneffectivedistributedratecontrolschemeisneededfortheadhocmode. lenTsuc;(4.8) whereUisthemappingfunctionfromtrafcratetochannelutilization,andTsucisdenedinequation( 4.1 )or( 4.2 ).Thus(cu,cupeak)specifyaow'sbandwidthrequirement,wherecu=U(TR)andcupeak=U(TRpeak). Onthesideofthecoordinator,thetotalbandwidthoccupiedbyalladmittedreal-timeowsisrecordedintwoparameters,i.e.,theaggregate(cu,cupeak),denotedby(cuA,cupeakA),whichareupdatedwhenareal-timeowjoinsorleavesthroughthefollowingadmissionprocedure.

PAGE 94

Whenreceivingareal-timeconnectionrequestfromitsapplicationlayer,anodemustsendarequestwithspecied(cu,cupeak)tothecoordinator,notingthatitwantstoestablishareal-timeow.Onlyaftertherequestisadmitted,thenodestartstoestablishtheowwiththeintendeddestination.Otherwise,thenoderejectstherequestandinformsthecorrespondingapplication. UponreceivingaQoSrequestwithparameters(cu,cupeak),thecoordinatorcheckswhethertheremainderofthequotaBMcanaccommodatethenewreal-timeow.Speci-cally,itcarriesoutthefollowing: Finally,whenareal-timeowends,thesourcenodeoftheowshouldsendacon-nectionterminatedmessagetothecoordinator,andthelatterrespondswithaterminationconrmedmessageandupdates(cuA,cupeakA)accordingly. Notethatreal-timepacketshavehighestpriorityintheoutgoingqueue,whichmeanstheywillalwaysbeputonthetopofthequeue.Meanwhile,allthecontrolmessagesrelatedtoconnectionadmissionandterminationaretransmittedasbestefforttrafc;however,theyhavehigherprioritythanotherordinarybesteffortpackets,whichhavethelowestpriority.Bydoingso,wemakesurethatthesemessagesdonotaffectthereal-timetrafcwhilebeingtransmittedpromptly. 79 ].Wewillfurtherinvestigatethisissueinourfuturework.

PAGE 95

wherekistheslidingwindowsize.Thustheinstantaneousavailablebandwidthforbestefforttrafc,denotedbycubi,is IftherecentkpacketsareallTCPpackets,thencuAri=0andallthebandwidthwillbeallocatedtoTCPows.Onceareal-timepacketwhichhashigherpriorityintheoutgoingqueueistransmittedorreceived,therateofTCPowswillbedecreased.ThisalgorithmthuseffectivelyadaptsTCPratetothechangeofVBRtrafcrate.Clearly,ifkissmall,theestimationisaggressiveinincreasingTCPrate;ifkislarge,theestimationisconservative[ 79 ].Wesetkto10inoursimulationasatradeoff. Givencub,thetaskistofairlyallocatethebandwidthtoallthenodesthathavethebestefforttrafctotransmit.Weassumethenumberofnodesthatarethesourcesofdownstreamowsisnd,andthenumberofnodesthatarethesourcesofupstreamowsisnu.Obviously,theaccesspointknowsbothndandnu.ThusthetrafcrateforthebestefforttrafcallocatedtotheaccesspointTRbaandthatallocatedtoeachmobilenodeTRbmareasfollows.

PAGE 96

whereU1istheinversefunctionofUdenedinEquation( 4.8 ). ThisrateallocationTRbaimmediatelytakeseffectattheaccesspoint.AndtherateallocationTRbmispiggybackedtoeachmobilenodebyusingtheMAClayerACKframeforeachbesteffortpacketfromthenode.Inthisway,themobilenodecanimmediatelyadjustthetransmissionrateofitsownbestefforttrafc.TwobytesneedtobeaddedintheACKframetoindicateTRbmwithaunitofRD216,whereRDisthebitratefortheMAClayerDATApackets. Notethattheabovefairallocationalgorithmisonlyonechoiceforratecontrol.De-pendingontrafcpatterns,otherallocationalgorithmscanalsobeused,sincetheaccesspointcanmonitortheinstantaneousrateofeachbesteffortowsfrom/toeachmobilenode.Forinstance,itiseasytodesignanalgorithmthatallocatesdifferentratetodifferentowsbymodifyingEquation( 4.11 ).Ratecontrolinadhocmode whereTRbnewandTRboldarethevalueofTRbafterandbeforetheadjustment,andRbtisathresholdofchannelbusynessratioandissetto95%BU.TwopointsarenotedonEquation( 4.12 ).First,weseethatthenodeincreasestherateofbestefforttrafcifRb
PAGE 97

wherePTRboldU1(RbRbr)isduetothefactthatRsRbasshowninEquation( 4.7 )andRbRbristhecontributionfromthetotalbestefforttrafctoRb.ThusafteronecontrolintervalTrb,thechannelutilizationwillapproximatelyamounttoRbt. ApparentlythisschemedependsontheestimationofRbr.LargerestimateofRbrresultsinlargerincreaseintrafcratewhenRbt>RbandlargerdecreaseintrafcratewhenRbtRbandsmallerdecreaseintrafcratewhenRbt
PAGE 98

Toenforceaconservativelyincreasingandaggressivelydecreasinglaw,wethussetRbrasfollows: WealsoneedtodeterminethecontrolintervalTrbdistributedly.Toberesponsivetothechangeofthechannelbusynessratioobservedintheair,therateisadjustedateachtimeinstantwhenanodesuccessfullytransmitsabesteffortpacket.ThusTrbissettotheintervalbetweentwosuccessivebesteffortpacketsthataresuccessfullytransmitted.Notethatevenwhensuchanintervalisshortandthusnoreal-timetrafcisobservedinit,i.e.,Rbr=0,therateofbestefforttrafccanatmostbeincreasedtoU1(Rbt).Atthattime,thecollisionprobabilityisstillverysmallaccordingtopreviousanalysis,sothereal-timepacketslateroncanbequicklytransmitted,whichwillinturnlowerthebestefforttrafcrate. 106 ].Inthissection,weevaluateitseffectivenessinan802.11wirelessLAN. 4 Tomodelmultimediatrafc,threedifferentclassesoftrafcareconsidered:

PAGE 99

Duringsimulation,theRTS/CTSmechanismisusedforvideoandTCPpackets,butnotusedforvoicepacketsbecauseofitsrelativelylargeoverhead.Thetrafcloadisgrad-uallyincreased,i.e.,anewvoice,videoorTCPowisperiodicallyaddedinaninterleavedway,toobservehowCARCworksandtheeffectofanewlyadmittedowontheper-formanceofpreviouslyadmittedows.Specically,until95second,anewvoiceowisaddedatthetimeinstantof6isecond(06i615).Likewise,avideoowisaddedtwosecondslaterandaTCPowisadded4secondslater.Furthermore,tosimulatetherealscenariowherethestartofreal-timeowsarerandomlyspreadovertime,thestartofavoiceowisdelayedarandomperioduniformlydistributedin[0ms,40ms],andthatofavideoowdelayedarandomperioduniformlydistributedin[0ms,125ms].Notethatinthesimulationperiodbetween[95ms,120ms],wepurposelystopinjectingmoreowsintothenetworkinordertoobservehowwellCARCperformsinasteadystate. Twoscenariosshownbelowareinvestigated.

PAGE 100

AscanbecalculatedusingEquation( 4.8 ),eachvoiceowcontributes0:0347tothechannelbusynessratioRb,andeachvideoow0:04339bynoticingthateachpacketisaddeda20bytesIPheaderinns-2.Thusafter12voiceand11videoowsareadmitted,theportionofRbthataccountsforthevoiceowsis00:38,withameanof0:19,andtheportionthataccountsforthevideoconnectionsis0:52.ThusU(TRA)=0:71,andU(TRApeak)=0:90.Thereafter,theadmissioncontrolmechanismstartstorejectfuturereal-timeows.Infrastructuremode 4(a) showsthethroughputforthethreetrafctypesthroughoutthesimulation.Atthebeginning,theTCPtrafchashighthroughput;thenasmorereal-timeowsareadmitted,itgraduallydropsasaresultofratecontrol.BecausewesetanupperboundBMforreal-timetrafc,itcanbeobservedthatwhenthetrafcloadbecomesheavy,TCPtrafc,asdesired,isnotcompletelystarved.BecauseTCPtrafcisallowedtouseanyavailablechannelcapacityleftbyreal-timetrafc,thetotalchannelthroughput,namelythesumofthethroughputduetodifferenttypesoftrafc,alwaysremainssteadilyhigh.NotethatthethroughputfortheTCPtrafcdoesnotincludethecontributionfromTCPACKpackets,eventhoughtheyalsoconsumechannelbandwidthtogetthrough.Thus,thetotalchannelthroughputshouldbesomewhathigherthanthetotalthroughputasshowninFig. 4(a) Theend-to-enddelayisillustratedinFig. 4(b) ,inwhicheverypointisaveragedover2seconds.Itcanbeobservedthatthedelayforreal-timetrafcisalwayskeptbelow20ms.Initially,asthenumberofadmittedreal-timeowsincreases,thedelayincreases.NotethattheincreaseofdelayisnotduetoTCPtrafc,butduetotheincreasingnumberofcompetingreal-timeows.Then,thedelayoscillatesaroundastablevalue.Fig. 43(c) presentsthedelaydistributionforvoiceandvideotrafc.MoredetailedstatisticsofdelayanddelayvariationaregiveninTable 4 andFig. 4 .AsshowninTable 4 ,the97percentiledelayvalueforvoiceandvideois35.5msand32.2msrespectively,and

PAGE 101

(b)Averageend-to-enddelayofvoiceandvideotrafc (c)End-to-enddelaydistributionofvoiceandvideotrafc Infrastructuremode:thenumberofreal-timeandTCPowsincreasesovertime.Channelrateis2Mbps. the99percentiledelayvalueforvoiceandvideois55.4msand45.2msrespectively.Itisknownthatforreal-timetrafc,packetsthatfailtoarriveintimeissimplydiscarded.Giventheallowable1%3%packetlossrate,thesedelaysarewellwithintheboundsgiveninSection 4.2.2 .ThegooddelayperformanceindicatesthattheCARCschemecaneffectivelyguaranteethedelayanddelayjitterrequirementsofreal-timetrafc,eveninthepresenceofhighlydynamicTCPtrafc. Finally,wenotethatinsimulation,nolostreal-timepacketisobserved.ThisshouldbeaccreditedtothefactthatourCARCschemesuccessfullymaintainsaverylowcollision

PAGE 102

Figure4: End-to-enddelayofallvoiceandvideopacketsininfrastructuremode Table4: Themean,standarddeviation(SD),and97'th,99'th,99.9'thpercentiledelays(inseconds)forvoiceandvideointheinfrastructuremode. SD 97%ile 99%ile 99.9%ile VBRVoice 0.0097 0.0089 0.0306 0.0412 0.0670 CBRVideo 0.0127 0.0081 0.0314 0.0392 0.0609 4 illustratestheperformanceoftheCARCschemewhenitworksintheadhocmode.Again,theperformanceisverygood.TheCARCschemedeliversalmostthesamethroughputandaverageend-to-enddelay,andalsonolostreal-timepacketisobserved.However,asseenfromFig. 4(c) ,thedelayvariationisslightlylarger,whichisalsoconrmedinTable 4 andFig. 4 .Thisisduetotheimperfectestimationoftherateofreal-timetrafcintheadhocmode,aseachnodelocallyestimatestherate. Fig. 4 demonstratesthattheratecontrolschemeachievesastableandhighchannelutilization,i.e.,around90%,whenthenumberofvoice,video,TCPowsoractivenodesvariesandthepacketsizefordifferenttypesoftrafcisdifferent.Thechannelutilizationiscalculatedbysummingupallthecontributionofthevoice,video,TCPDATAandTCPACKpacketstothechannelutilizationaccordingtotheend-to-enddatarateasshowninFig. 4(a) andequation 4.8

PAGE 103

(b)Averageend-to-enddelayofvoiceandvideotrafc (c)End-to-enddelaydistributionofvoiceandvideotrafc Adhocmode:thenumberofreal-timeandTCPowsincreasesovertime.Channelrateis2Mbps. Thus,ourratecontrolschemeforadhocmodeprovidesanotherkindofdistrib-utedsolutiontomaximizingthenetworkthroughputbesidesthemethodsinthepapers[ 12 13 16 20 85 90 ].However,unlikethesepreviousapproaches,oursdoesnotchangethemediaaccessmechanisminDCFprotocolandhasastableperformanceunderdifferentnumberofactivenodesanddifferentpacketsizeinthepresenceofCBR,VBRandTCPbestefforttrafc.

PAGE 104

Figure4: End-to-enddelayofallvoiceandvideopacketsinadhocmode Table4: Themean,standarddeviation(SD),and97'th,99'th,99.9'thpercentiledelays(inseconds)forvoiceandvideointheadhocmode. SD 97%ile 99%ile 99.9%ile VBRVoice 0.0101 0.0104 0.0350 0.0500 0.0876 CBRVideo 0.0133 0.0092 0.0337 0.0477 0.0903

PAGE 105

Figure4: Channelutilizationinadhocmode layer.Andthecalladmissioncontrolschemecouldalsoconsiderretxwhenissuesthead-missions.ThusthewholeCARCschemecouldeffectivelysuppressedtheadverseeffortscausedbychannelfadingandstilldeliveracomparableQoSperformance. ItisimportanttonotethatnormallychannelfadingisnotaseriousproblemintheWLAN,whichfeatureslownodemobilityandrelativelystablechannel.However,ifthepacketerrorprobabilityduetochannelfadingbecomessignicant,theQoSlevelwillbehurt.However,ourproposedCARC,byconsideringtheretx,canstilleffectivelycontrolthetotalinputtrafcrateandhencemaintainaverysmallcollisionprobabilitytoguaranteethe802.11MACprovidesthebestQoSlevelitcansupportinthiscase.Ofcourse,ifchannelfadingisseriousenough,thisbestQoSlevelmaynotsatisfytheQoSrequirementofreal-timetrafc.

PAGE 106

orbestefforttrafc,nodifferentiationiscommitted.Asaresult,allthereal-timetraf-c,includingCBRandVBRtrafc,equallysharesthedelayanddelayvariation,whichsometimesisnotexibleenough. Ifaprioritized802.11MACprotocolsimilartotheschemes[ 1 125 ]isadopted,weareabletoprovideprioritywithinreal-timetrafc.Asaresult,thehighpriorityreal-timetrafcreceivessmallerdelayvariation,whereasthelowpriorityreal-timetrafcreceiveshigherdelayvariation[ 33 ].Ofcourse,tofullyexploitthepotentialoftheprioritizedMACandmeetdifferentQoSrequirements,theadmissioncontrolandratecontrolalgorithmsproposedhereshouldcontroltheaggregaterateofeachclassoftrafcsothatcollisionswithineachclassissmallenoughtoguaranteethatitsQoSrequirementisnotviolated. 150 ],inthischapterwehaveproposedasim-pleandeffectivecalladmissioncontrolandratecontrolscheme(CARC)tosupportQoSofreal-timeandstreamingtrafcinthe802.11wirelessLAN.Basedonthenoveluseofthechannelbusynessratio,whichisshowntobeabletocharacterizethenetworkstatus,theschemeenablesthenetworktoworkattheoptimalpoint.Consequently,itstatisti-callyguaranteesstringentQoSrequirementsofreal-timeservices,whileapproachingthemaximumchannelutilization. Furthermore,theratecontrolschemeforadhocmodehasitsownvirtue.Itprovideanotherkindofdistributedsolution,i.e.,ratecontroloverthepacketsinoutgoingqueuewithoutmodicationtothemediumaccessmechanismintheIEEE802.11DCFprotocol,tomaximizethenetworkthroughput,andhasstableperformanceunderdifferentnumberofactivenodesanddifferentpacketsizeinthepresenceofalltheCBR,VBRandTCPtrafc. Combiningtheanalyticalresultsinourpreviouswork[ 150 ]andourproposedCARCscheme,wethereforemakeitclearthattheIEEE802.11WLANcanprovidestatisticalQoSguarantees,notjustdifferentiatedservice,formultimediaservices.

PAGE 107

Recentyearshaveseenincreasinglygrowingusersinwirelesslocalareanetworks.Tobettermeetuserneeds,varioustypesofapplicationsincludingvoiceoverIP(VoIP),streamingmultimedia,anddataservicesareexpectedtobesupported.However,theIEEE802.11distributedcoordinationfunction(DCF)standardhaslongbeenknownasinefcientandunfairinthepresenceofmanyconcurrentusers.Furthermore,despitetheavailabilityofmanyservicedifferentiatedschemes,qualityofservice(QoS)forreal-timeservicesisstillnotwellsupported. Toaddresstheefciency,fairnessandQoS,weproposeadistributedresourcealloca-tion(DRA)frameworkontopof802.11.DRAreliesonthenoveluseofchannelbusynessratio(BR)asthenetworkstatusindicator,whichcanbeeasilyobtainedin802.11.BasedonBR,anovelthree-phasecontrolmechanismisproposedtofairlyandefcientlyutilizenetworkresourceandguaranteeashortmediumaccessdelay.DRAalsointegratesthethree-phasecontrolmechanismwithacalladmissioncontrolschemeandapacketconcate-nationschemeintoasingleuniedframeworktobettersupportQoSandmultiplechannelratesbesidestheefciencyandfairness.ExtensivesimulationsdemonstratethatDRAachievesnearoptimumthroughputandtheperformanceisstableregardlessofthenumberofnodesandpacketsizes.Comparedto802.11,itimprovesthroughputbyashighas71%withRTS/CTSand157%withoutRTS/CTS.Fairnessiswellachieved.Moreover,DRAcanprovidestatisticalQoSguaranteeforreal-timeservices. 68 ]basedwirelesslocalareanetworks(WLANs)havebeenwidelydeployedduetotheirlowcostandeasyaccessibility.Distributedcoordination 88

PAGE 108

function(DCF),thefundamentalchannelaccessmethodinthe802.11MAC,isbasedoncarriersensemultipleaccesswithcollisionavoidance(CSMA/CA).Brieyspeaking,itusesabinaryexponentialbackoff(BEB)schemetoreducethecollisionprobabilityandtheRTS/CTSexchangebeforedatatransmissiontoshortenthecollisionperiodsoflongdatapackets.Whileitworkswellinsupportingtraditionalbest-efforttrafc,itisinadequateindealingwithseveralcriticalissuesthatariseasaresultoftheever-increasingnumberofusersandtheirdiverseserviceneeds. First,althoughthedatarateofWLANshasincreaseddramatically,itisstillbelievedtolagbehindtheincreasingbandwidthdemands.Consequently,thenetworkseasilyentersaturation.Inaddition,duetothepopularuseofWLANtechnology,itisnotuncommontoseealargenumberofconcurrentusersaccessingtheWLANthroughthesameaccesspoint(AP),suchasinconferencehallsorclassrooms.However,theinherentdeciencyofDCFinsupportingmanyusersconcurrentlyunderheavytrafcloadalwaysresultsinseverepacketcollisionsandhencegreatlydegradesnetworkthroughput([ 21 85 160 15 ]).Therefore,efcientlycoordinatingsimultaneouschannelaccessbymanyusersisimpor-tant. Second,networkusagethatusedtobedominatedbywebtrafchasshifteddramati-callywithsignicantincreasesinVoIP,streamingmultimediatrafcandpeer-to-peertrafc[ 62 ].ComingtogetherwiththisshiftisthedemandforQoSsupportinWLANs.However,supportingQoSinwirelesschannelsisdifcultgivenanumberofchallenges[ 125 ].Be-sidesthewirelesschannelerrors,thenetworkthroughputvarieswiththelevelofchannelcollision;sodoesthepacketdelay. Third,giventheconictbetweenthelargenumberofusersandtherelativelylimitedchannelcapacityandthediverseQoSrequirementsimposedbyarangeofapplications,fairlyallocatingchannelresourcetoallusersishighlydesirable.Whilemanygoodfairqueueingschemeshavebeenproposedforthewirednetworksandcellularnetworks[ 166 97 ],theymaynotbedirectlyapplicabletoWLANsifthenetworksoperateinadhoc

PAGE 109

mode.Inthiscase,duetothelackofacentralizedcontroller,nonodeinthenetworkhasaglobalviewofthenetworkstatus,suchasthenumberofcontendingnodesorthetrafcsituationateachindependentnode.Clearly,thisdictatesadistributedfaircontrolmechanism.Furthermore,adistributedmechanismhasseveralbenetscomparedtoacentralizedone,suchastheavoidanceofsinglepoint-of-failureandscalability. Intheliterature,enormouseffortshavebeenspentindealingwitheachoftheaboveissues.Althoughsignicantprogresshasbeenmade,noneofthemiscompletelyaddressed(moredetailsaregiveninSection 5.6 ).Whilethephilosophyofbreakingupabigproblemintoseveralsmalleronesandaddressingeachindividuallyiscommonandeffective,ajointandsystematicstudyoftheseissuesmayleadtoagooduniedsolutionforthefollowingreasons.

PAGE 110

Motivatedbythosepotentialbenets,weendeavortoseekauniedsolutiontothethreeissues.Tothebestofourknowledge,therehasbeennosuchstudythusfar.More-over,consideringthelimitedprocessingpower/batterypowerateachmobilenodeandthedistributednatureofDCF,itisdesirablethatanysolutionaimedatenhancingDCFissimpleandfullydistributed.Infact,simplicityanddistributedcontrolaretheverycharac-teristicsthatcontributetotheunprecedentedsuccessofWLANs.Also,thesolutionwithasfewaspossiblemodicationsto802.11ispreferredforthepurposeofcompatibility. Inthischapter,weconductacomprehensivestudyoftheissuesofefciency,fairnessandQoSandproposeanoveldistributedresourceallocation(DRA)frameworkfor802.11WLANs,whichisfair,efcient,andQoS-capable. Thecontributionofthischapteristwofold.First,usingthechannelbusynessra-tiotocharacterizethenetworkstatus,wedevelopanovelthree-phasecontrolmechanismtodynamicallyallocatenetworkresourceintermsofchanneltime.Bywellcontrollingchannelcollisions,thismechanismachievesnearoptimumthroughput,fairresourceallo-cation,andsmallMACdelay.Theoreticalanalysisshowsthatthemultiplicative-increasephasequicklyleadstotheconvergencetohighefciency,andtheadditive-increaseandmultiplicative-decreasephasesquicklyleadtotheconvergencetofairness.Second,tobet-tersupportQoSandmultiplechannelrates,wefurtherproposeauniedframework,DRA,whichincorporatesthethree-phasecontrolmechanismandseveralothermechanisms.Byconductingcalladmissioncontroloverreal-timeservicesandproperlyadjustingthenet-workresourceamongreal-timeandnon-real-timeservices,DRAcansupportstatisticalQoSforreal-timeserviceswithshortdelayandzeropacketlossrate.InDRA,wealsode-velopapacketconcatenationschemetoreducetherelativelyhighcontroloverheadwhenmultiplechannelratescoexist.ExtensivesimulationsverifytheperformanceofDRA.Insummary,DRAhasthefollowingdesirablefeaturesthatdistinguishitselffrompreviousschemes:

PAGE 111

Therestofthischapterisorganizedasfollows.ThedesignrationaleforDRAisgiveninSection 5.2 .WedescribeDRAindetailinSection 5.3 .TheconvergenceanalysisispresentedinSection 5.4 .TheperformanceevaluationisconductedinSection 5.5 .Section 5.6 discussestherelatedwork.Finally,Section 5.7 concludesthischapter. 5.2.1EfciencyandQoS ForDCF,thenetworkthroughputScanbeexpressedastheaveragepayloadsizetransmittedinatimeslotdividedbytheaveragelengthofatimeslot.Then,followingthetechniquesofBianchi'spaper[ 15 ],wecanderivethethroughput: (1Ptr)+PtrPsTs+Ptr(1Ps)TcPtr=1(1)nPs=n(1)n1 InEquation( 5.1 ),E[D]istheexpectedpayloadsize,Tsistheaveragesuccessfultrans-missiontime,Tcistheaveragecollisiontime,isaMAClayeridleslottime,isthe

PAGE 112

transmissionprobabilityofeachnodeinanyslot,nisthetotalnumberofnodesintheWLAN,andpisthecollisionprobabilitythatanodeencounterscollisionwhenevertrans-mitting. Focusingonthesaturatedcasewhereeachnodealwayshaspacketsinitsqueueawait-ingtransmissions,Bianchiderivedtheformulafor, (12p)(CWmin+1)+pCWmin[1(2p)m]m=logCWmax whereCWminandCWmaxaretheinitialandmaximumcontentionwindow.ByEquations( 5.1 )and( 5.2 ),,pandScanbesolved. However,themaximumthroughputisnotnecessarilyachievedinthesaturatedcase.Denotebyp0andS0thevaluesofpandSforthesaturatedcase,thentheactualcollisionprobabilitycouldbelessthanp0ifnotallthenodesinthenetworkarecontendingforthechannelatthesametime,potentiallyleadingtoathroughputhigherthanS0.NotetheexpressionofSinEquation( 5.1 )isgeneralenoughandcanbeappliedtothenon-saturatedcaseaswell.ToobtainthemaximumvalueofS,denotedbyS,andthecorrespondingvalueofp,denotedbyp,wecanrewriteSasafunctionofpandlet dpS=0(5.3) Further,wedenotebyprtherootofEquation( 5.3 ).Sincegivenn,pisupperboundedbyp0,weobtainpas BoththemaximumandsaturatedthroughputareplottedinFig. 5 usingthesameformu-lasofTsandTcasusedbyBianchi[ 15 ]. Itisclearthatthemaximumthroughputcannotbeachievedinthesaturatedcasees-peciallywhennisnotsmall.Itisimportanttonotethatthemaximumthroughputisnotsensitivetothenumberofnodes.Equation( 5.4 )impliesthatsomehowifwecantunep

PAGE 113

Figure5: Maximumandsaturatedthroughputwithdifferentnumberofnodes(RTS/CTSisused,packetlength=1000bytes,channelrate=11Mbps) toapproachp,thenetworkcanattainthemaximumthroughput.However,sincepisnoteasilyobtainableandcontrollable,weareforcedtoseekagoodalternative.Letbrdenotethechannelbusynessratio,i.e.,theratioofthetimewhenthechannelisbusy,anditcanbeexpressedas Itcanbetriviallyshownthatbrisaninjectivefunctionofp;moreover,itcanbeeasilymeasuredsince802.11isbasedoncarriersensing.Denotebybrthecorrespondingvalueofbrwhenp=p,wethuscantunethenetworktoworkatbr. Recently,we[ 150 ]havetheoreticallyshownthatbrisrelativelystableandaround0:900:98,andifbr6br,thedelayisgoodenoughtosupportstatisticalQoS.ItisthusreasonabletomakethenetworkdeliverhighthroughputandsupportQoSbyallowingbrtoapproachbrandensuringbr6br.

PAGE 114

First,throughputunfairnessmightariseifnodesusedifferentpacketsize,whichseemsin-evitableasthesupportedservicesinWLANsbecomediversied.Moreover,theaggregatethroughputwillalsobehurtifsomenodesuseaverysmallpacketlengthandhenceintro-ducelargecontroloverhead.Second,sincethebackoffprocessinDCFalwaysfavorsthosenodesthatjustsuccessfullytransmittedapacket,severethroughputunfairnessmighthap-peninshortperiods.Third,as802.11supportsmultiplechannelbitrates,ifdifferentnodesusedifferentrates,thisapproachwillpenalizenodeswithhighratesandsignicantlylowertheaggregatethroughput,asrevealedbyHeusseetc.[ 63 ].Therefore,weaimtoachievetimefairness[ 121 ].Specically,DRAisdesignedtoefcientlyandfairlyallocatethechanneltimeamongnodes.Thistimefairnessmodelcanavoidtheaboveproblemswiththethroughputfairnessmodel.Inaddition,itcanstrikeagoodbalancebetweenefciencyandfairnessduetovaringchannelqualities.Withequallyallocatedchanneltime,nodeswithbetterchannelqualitycangethigherthroughputwithhigherphysicalchannelbitrate. Atrst,wedescribethebasicframework.Inthisframework,itisassumedthatthereexistsacalladmissioncontrollerensuringthattheadmittedreal-timetrafcislessthanthenetworkcapacityinordertosupportQoS.Therefore,inthefollowingdescription,werstfocusonhowtoadjustthesendingrateofnon-real-timetrafctoachievehighefciency,goodfairnessandshortMACdelay. ThenweexplainhowDRAsupportsQoSandmultiplechannelratesbyincorporat-ingnecessarycomponentssuchascalladmissioncontrol,priorityqueue,andthechanneladaptivepacketconcatenationmechanism.

PAGE 115

Wedenechanneltimeforpacketp,denotedbytp,asthetimethatasuccessfultransmissionofpacketpwilllastoverthechannel.Accordingtothe802.11standard,wethushave forthecasewheretheRTS/CTSmechanismisused,and forthecasewherethereisnoRTS/CTSmechanism.InEquations( 5.6 )and( 5.7 ),rts,cts,data,andackarethecorrespondingtransmissiontimesforMACframesRTS,CTS,DATAandACK,respectively;sifsanddifsarethemandatoryinter-framespacesbetweentheseframes.Notethatdifsisincludedintpbecauseeachnodeisrequiredtoobservethechannelidleforatleastdifslongbeforebackingofforstartingnewtransmissions. Aftercalculatingtp,DRAcanobtainthescheduledinterval,thetimebetweentwoconsecutivepacketsthatDRApassestotheMAClayer: Inthisway,thetrafcrateofthenon-real-timetrafcateachnodeisdetermined.Notethatboththepacketlengthandchannelbitrateareguredin.

PAGE 116

whereuisarandomnumberuniformlydistributedin[0;1].Thisrandomdelayisusedtoavoidthefollowingundesirablesituation.Whenmanynodesjointhenetworkatthesametime,theymayallobservetheidlechannelandstartcontendingforthechannelsimultaneously,therebycausingseverecollision.InstantaneousRateUpdateProcedure MAC callback time(5.10) wheretnowisthecurrenttime,tlast MAC callback timeistheMACcallbacktimeforthelastpackettransmission,andtpisthechanneloccupationtimereturnedbytheMACcallbackfunction.NotethattpcanalsobecalculatedbasedonthecurrentchannelbitrateaccordingtoEquations( 5.6 )and( 5.7 ). DRAregardsrintasthecurrentallowableresourceandcalculatesanewallowableresource,denotedbyrnew,withthefollowingthree-phaseresourceallocationmechanism.Three-PhaseResourceAllocationMechanism

PAGE 117

Intheabovephases,bristhecurrentchannelbusynessratiocontributedbyallthenodesinthenetwork;brthisthechannelbusynessratiothatisaconstantveryclosetobr;BMisathresholdthatdetermineswherethechannelstatuschangesfromunderloadedtomoderatelyloaded. Noticethesummationofrintoverallnodesisequaltobr.IfeachnodeincreasestheallowableresourcebyEquation( 5.11 ),thechannelbusynessratiowillquicklyconvergetobrthafteraperiodduringwhicheachnodetransmitsonemorepacketonaverage.Ifsomenodesnolongerincreasetheirtrafcratesduetotheconstraintsoftheirapplications,othergreedynodes Wheneverbr>BM,DRAadoptsanadditive-increaseandmultiplicativedecrease(AIMD)algorithmtoconvergetobothhighefciencyandfairness. whereistheincreaseparameter.

PAGE 118

whereisadecreaseparameter,and0<61. Thevaluesofandimpacttheconvergencespeedofboththeefciencyandfairness.WewilldiscusshowtosettheseparametersinSection 5.4 .BackoffProcedure sending time.Ifitislargerthanbrth,DRAwillscheduleanadditionaldelaydbeforepassingapackettotheMAClayer: sending time)(5.14) wheretnowisthecurrenttime.Otherwise,DRAwillimmediatelysendthepackettotheMAClayer.AcquisitionofChannelBusynessRatio Thereisalreadyafunctioninthe802.11MACtodeterminewhetherthechannelisbusyornot.Thechannelisconsideredbusywheneverthenodeunderconsiderationistransmittingorreceiving,orphysicalcarriersensingornetworkallocationvector(NAV)indicatesabusychannel.Thechannelbusynessratiocanbecalculatedbyaddingupallthebusyperiodsandthendividingthesumbytheobservationperiod.DRAcandeterminethestartandendpointsoftheobservationperiod.ItthuscanbeseenthattheacquisitionofchannelbusynessratioonlyrequiresseveralsimplecalculationsattheMAClayer.

PAGE 119

11 ]meansthat,foreachsessionp,theraterpcannotbeincreasedwithoutdecreasingtherateforsomesessionp0forwhichrp06rp.Itisachievedwheneachsessionhasabottlenecklink. IntheWLAN,allnodessharethesamewirelesschannelandeachreachesitsdestina-tionviaonehop.Toestablishthemax-minfairness,onemayregardtheapplicationlayerasthebottlenecklinkifitinjectslesstrafcthantheMACcantransmitinthefairshareofthechanneltime.DRAcanachievethenewmax-minfairnessobjectivebysuccessfullytransmittingallthepacketsfromthenodeswhichhavethebottlenecklinkattheapplicationlayer.Andalsoitallowsothergreedynodeswhosebottlenecklinkisthesharedwirelesschanneltofairlysharealltheresidualchannelresource. Incontrast,intheoriginal802.11WLAN,packetcollisioncouldbesevereifmanynodesareheavilyloaded.Thismayleadtopacketdropsintwoways.First,accordingtothe802.11standard,packetscouldbedroppedduetoconsecutiveretransmissionfailures.Second,highcollisionprobabilitymaycauseexcessivelylargemediumaccessdelayandhencethebuildupofthequeue.Thenpacketswillbediscardedifthequeueisfull.802.11itselfmayfailtoachievemax-minfairnessbydroppingpacketsfrombothgreedyandnon-greedynodes.

PAGE 120

forthenetworktosupportshortdelayanddelayvariationasrequiredbyreal-timetrafc,thisisnotsufcientforthefollowingreason.ThedelayapacketexperiencesinaWLANsconsistsofthequeueingdelayandthemediumaccessdelay.Eventhoughthelattercanbewellcontrolled,theformerwillbeexcessivelylargeifthetotaltrafcloadofreal-timetrafcexceedsthenetworkcapacity.Therefore,calladmissioncontroloverreal-timetrafcisneeded. Inthischapter,wedonotdetailanadmissioncontrolalgorithmduetospacelimitation.However,somecalladmissioncontrolschemesrecentlyproposedfortheWLANssuchas[ 149 ]canbeused.PriorityQueueSchemeforAdmittedReal-TimeTrafc

PAGE 121

5.4.2 ,differentchannelbitratesdonotimpacttheconver-genceoftimefairness.Inotherwords,evennodesusedifferentchannelrates,DRAfairlyallocatesthechanneltimeamongthem.Asaresult,nodeswithhigherchannelrateswillbeabletotransmitmorepacketsinthesametimeperiodthanthosewithlowerchannelrates,whichtranslatesintohigherthroughput.Clearly,DRApreventsthedegradationofaggregatethroughputthatoccursinmulti-rate802.11.However,inthisapproach,evenwhenthechannelrateishigh,onlyonepacketcanbetransmittedinoneDATA/ACKorRTS/CTS/DATA/ACKhandshake.Ifthepacketsizeissetaccordingtothebaserateormediumrate,unnecessarycontroloverheadisintroduced. Tofurtherreducetheoverheadandimprovethroughputintheallocatedchanneltime,DRAadoptsachanneladaptivepacketconcatenation(CAPC)mechanism.Whenthechan-nelrateishighduetoagoodchannelcondition,sincechannelcoherencetimetypicallyex-ceedsmultiplepackettransmissiontimes[ 113 ],DRAcanconcatenateseveralshortpacketsintoonelargepacketforMAClayertransmission.Thenumberofpacketsboundinasingletransmissioncanbeashighastheratiobetweenthecurrenthighchannelrateandthebaserate.Finally,wenotethatDRAcanworkwithsomeexistingrate-adaptiveschemes,suchas[ 65 113 76 ],toachievebetterperformance.

PAGE 122

multiplicative-increasephaseendsandthenetworkgoesintotheAIMDphases.Next,weshowhowlongittakesforDRAtoentertheAIMDphaseswhenagreedynodejoinsthenetworkandndsthechannelisunderloaded,i.e.,br1)istheupdatedresourceallocationaftertheithpackettransmission.Weseethatthisisaveryaggressiveincreasephase.Therstnodecanachievethehighestallowablecapacityaftertherstpackettransmission.Case2: Theorem1 brth(brthBM)rstart

PAGE 123

Similarly, wherernew0isequaltorint,theinstantaneouscalculatedresourceconsumedbytherstpackettransmission.Whenrnewn+rb>BM,i.e.,br>BM,thenetworkenterstheadditive-increasephase.Tocalculatethesmallestnnecessaryforthenetworktoentertheadditive-increasephase,wesolveforthesmallestnsuchthat Denotesuchnbyn0.Then,wegetthecorrespondingtime0MIas Apparently,n0and0MIarebothdecreasingfunctionsofrnew0.FromSection 5.3.1 ,weknowthatrnew0>rstart,thenwecanobtaintheirrespectiveupper-bounds,denotedbyn

PAGE 124

Figure5: Convergencespeedofmultiplicative-increasephase(packetlength=1000bytes,channelrate=11Mbps) andMIbyreplacingrnew0withrstart.LetequalityholdinEquation( 5.25 ),weget brth(brthBM)rstart wherefunctiondxeisthesmallestintegerthatisgreaterthanorequaltox.Theorem1fol-lowsfromEquations( 5.24 ),( 5.27 )and( 5.28 ). 5 .Itcanbeseenthatthenetworkwillentertheadditive-increasephaseafterthenewnodenishestransmissionsfor1to31packets,orequivalently0:24:2s.Thesmallerrstartis,thelargerMIis.

PAGE 125

AlthoughNrcanbeanyvaluebetweenbrthandbrth,wewillshowlaterthattheinitialvalueofNrdoesnotmatterintheanalysis.Therefore,weassumeNr+rb+rstart=brthwhenanewnodewiththeinitiallyallocatedresourcerstartjoinsthenetwork.AccordingtoEquation( 5.13 ),thenthenewnode'sresourceischangedtornew0=rstart,andeachoftheNnodes'allocatedresourcesischangedtoanewvaluer0=r.Thetotalusedresourcebecomes(brthrb)+rb,andthetotalavailableresourceRais TheneachnodewillincreaseitsresourceaccordingtoEquation( 5.12 ). Wedenotebyincrease-decreaseperiodtheperiodfromthetimewhennodesbegintoincreasetheirresourcetothetimewhentheallocatedresourceisdecreasedbytheratio

PAGE 126

wheretpnewandtpkarethenewnode'sandnodek'schanneltransmissiontimeofonepacket,respectively.ByEquation( 5.35 ),weobtain Meanwhile,accordingtoEquation( 5.8 ),weknowtpnew forallkalthougheachnodemaytransmitadifferentnumberofpacketsintheithperiodduetodifferentpacketlengthorchannelbitrate.Attheendoftheithperiod,thetotalincreasedamountshouldbeequaltotheavailableresourceRa,i.e.,

PAGE 127

whichimpliesthatthevalueofTiisthesameforalli.Therefore,Equation( 5.37 )leadsto Whenthenetworkreachestheequilibrium, Thenallnodesincludingthenewnodewillhavethesameallocatedresourcewhichdy-namicallychangesbetween(brthrb) Next,letusderivetheconvergencespeed.Weconcludethatthenodesconvergetofairnessifthefollowingconditionismet: whereisarealnumbercloseto1,and>1ifrnew0>r0,<1ifrnew0
PAGE 128

Figure5: ConvergencespeedofAIMDphaseswhen=0:5 5 showstheconvergencetimeisverysmall.Andintuitively,iftheinitialresourcerstartisaroundthefairshare,theconvergencetimeisalmostequaltozero. 1 .OnceDRAisintheAIMDphases,nodek's(16k6N)resourcerkconvergestothenewfairsharer=brthrb

PAGE 129

Forhighefciency,shouldbecloseto1sincetheutilizedresourcedynamicallychangesbetweenbrthandbrth=BM.Weset=0:95inboththeanalysisandsimulationstudies.ItcanbeseenfromFig. 5 that=0:95canleadtoashortconvergencetime. FromTheorem 2 ,weseethatthefairnessconvergencetimeAIMDisinverselypro-portionalto.ToreduceAIMD,alargeisdesirable.Ontheotherhand,asmallispreferredtoreducethedegreeofoscillationineachnode'sallocatedresourceduringtheAIMDphases.Furthermore,wendintheextensivesimulationsthattheinstantaneousrateupdateandbackoffproceduresinDRAeffectivelyacceleratethefairnessconvergence.OnceanewnodejoinsthenetworkandbeginsthetransmissionattheMAClayer,othernodes'transmissionsaresloweddownduetoanobservedbusierchannel.Theseproce-duresalsohelpdampentheoscillationontheresourceadjustmentintheadditive-increasephasesincetheseproceduresalwaysreecttheinstantaneouslyachievablerate.Takingallthosefactorsintoaccount,weset=0:05.Thesimulationstudiesgivenbelowconrmitisagoodchoice. 5.5.1SimulationSetup

PAGE 130

Figure5: ImpactofpayloadsizeLandnumberofnodesnontheoptimalthresholdforchannelbusynessratiobrth Fig. 5 showsthatwithpacketlengthL=1000bytes,whenRTS/CTSisused,thethroughputofDRAismaximizedandalmostinsensitivetothenumberofnodesif

PAGE 131

MACdelayinDRAislessthan20msinalmostallthecaseswhenbrthislessthanorequalto0.95.Also,thecollisionprobabilityisaslowas0%20%ifbrthisproperlyset. ToavoidestimatingtheaveragepacketlengthtransmittedintheWLANs,weoptforasinglechannelbusynessratio.Wesetbrthas0:95and0:93respectivelyforthecasewithRTS/CTSandthecasewithoutRTS/CTSinallthefollowingsimulationstudies.Ifbothcasescoexistinthenetwork,werecommendthatbrthissetas0.93toguaranteeashortdelayandasmallcollisionprobabilityinallcases. Inthe500secondssimulationtime,anewnodejoinsthenetworkevery10seconds.Weobservetheinstantaneousthroughputofeachindividualnode.Fig. 5(a) showsthateachnodecanobtainafairresourceallocationwithin0to4secondsafteritjoinstheWLAN.AsshowninFig. 5(b) ,the802.11onlyperformswellinfairnessconvergencespeedwhenthenumberofnodesisverysmall,i.e.,lessthan10.The802.11isquiteunfairamongmorethan10nodesduringshortperiods,wheretheinstantaneousthroughputofeachnodeoscillatesinalargerange.Wealsoobservethattheaggregatethroughputofthe802.11dropswhileDRAmaintainshighandalmostunchangedthroughputasthenumberofnodesincreases. DRAalsosupportsmax-minfairness(refertoSection 5.3.2 )inthatitallowsallthetrafcfromthosenodeswhosetrafcrateislessthanthefairshareofthechannelresourcetogetthrough.Inthesimulationstudy,thereare10groupswith5nodeseach.Thenodes

PAGE 132

(b)802.11 FairnessconvergencewithRTS/CTS:onegreedynodejoinsthenetworkevery10seconds(packetlength=1000bytes,eachpointisaveragedover1second) ofeachgrouphavethesametrafcrate.Theratesforthe10groupsare0.2,0.4,0.8,1.2,1.6,2.0,2.4,2.8,3.2,and3.6Mbpsrespectively. Fig. 5(a) and 5(b) showthatDRAsuccessfullytransmitsallthepacketsofthosenodeswhosetrafcrateislowerthanthefairshareofthechannelresourcewithandwithoutRTS/CTS.Andweobserve04%packetlossesforthetherst15owsinthe802.11.Theselossesaremainlyduetothehighcollisionprobabilityin802.11.BothDRAandthe802.11willdroppacketsfromthoseowswithtoohighdatarates.Moreover,inthecaseofnoRTS/CTS,DRAcansupportthe5owsofthe1.2Mbpsdataratewithoutdropping

PAGE 133

(b)WithRTS/CTS packets.ThisisbecauseDRAachieveshigherthroughputincaseofnoRTS/CTSthanincaseofRTS/CTS,whichisalsoveriedinSection 5.5.4 Aggregatethroughputisimprovedby14.9%from4.54to5.22MbpswhenRTS/CTSisnotused,andby8.2%from3.85to4.17MbpswhenRTS/CTSisused.NoteDRAalsoachievesbetterfairnessforthoseowswithatrafcratelargerthanthefairshareofthechannelresource. Fig. 5 showsDRAcanprovidechanneltimefairnessunderdifferentchannelbitrates.Thereareatotalof30nodesinthenetwork,rst10nodesusing2Mbps,thesecond10nodesusing5.5Mbps,andthelast10nodesusing11Mbps.Thepacketsare1000byteslong.DRAstartsthechanneladaptivepacketconcatenationprocedureforthemultirate

PAGE 134

(b)Throughput DRA:fairnesswithmultiplechannelbitrates(RTS/CTSisused) (b)Throughput 802.11:fairnesswithmultiplechannelbitrates(RTS/CTSisused) WLAN.Itconcatenatesthreepacketsinonetransmissionforthe5.5Mbpsnodesand6packetsforthe11Mbpsnodes;itstilltransmitonepacketateachtimefor2Mbpsnodes.Incontrast,asthe802.11isdesignedtoachievethroughputfairness,slownodesusemuchmorechanneltimethanfastnodes,asillustratedinFig. 5 .Asaresult,theaggregatethroughputisgreatlyreduced.Comparedtothe802.11,DRAimprovestheaggregatethroughputby240%from1.3045Mbpsto4.4410Mbps. Fig. 5 (a),(b)and(c)showthatthroughputperformancestaysalmostthesameasthenumberofactivenodesincreasesinDRA,whereasitdegradesdramaticallyfor

PAGE 135

802.11.HeretheguresforthescenariowhereRTS/CTSisnotusedareomittedduetothesimilarity.Whenthenumberofactivenodesisequalto300,DRAimprovesthethroughputby71.62%,62.48%,59.22%,48.96%,36.63%withRTS/CTSand66.42%,84.01%,82.23%,126.86%,157.32%withoutRTS/CTSforpacketlengthsof200,500,1000,2000,5000bytesrespectively.Fig. 5 (d),(e)and(f)showthatDRAsupportsaveryshortmediumaccessdelayaslowaslessthan30ms,whichisdesirabletosupportthereal-timeservices.Thebadperformanceof802.11isattributedtothehighcollisionprobability,asconrmedinFig. 5 (g),(h)and(i). WealsocompareDRA'sthroughputwiththetheoreticalmaximumthroughputforthe802.11DCF,whichiscalculatedbypinEquation( 5.4 ).TheresultsshowthatthemaximumthroughputislargerthanDRAbyupto8.4%,5.52%,4.31%,5.04%,7.81%withRTS/CTSand10.16%,4.74%,7.35%,8.16%,12.22%withoutRTS/CTSforthepacketlengthsof200,500,1000,2000,5000bytesrespectively.ThedifferenceispartlyduetothechoiceofasinglechannelbusynessratioasshowninFig. 5 andispartlyduetotherequirementoffairnessadjustmentandshortMACdelayaswellasthedifcultytokeepchannelbusynessratioequaltotheoptimumvaluebr. 73 ],thetolerablepacketlossrateis13%andtheone-waytransmissiondelayispreferablyshorterthan150msbutshouldbenolargerthan400msforreal-timeservices. Inthissimulation,thereare100nodes.50nodeshavegreedytrafc.Fromsecond0on,every60secondsonenewnodestartsavideoow;fromsecond30on,every60secondsonenewnodestartsavoiceow.Allthesenodesrandomlychooseadestination. WeuseaCBRmodelforthevideoows.Therateissetas64kb/swithapacketsizeof1000bytes.Voicetrafcismodeledason/offtrafc.Theonandoffperiodsare

PAGE 136

(b)L=1000bytes (c)L=5000bytes (d)L=200bytes (e)L=1000bytes (f)L=5000bytes (g)L=200bytes (h)L=1000bytes (i)L=5000bytes Throughput,MACdelayandcollisionprobabilitywithRTS/CTS exponentiallydistributedwithanaveragevalueof300mseach.Duringtheonperiods,trafcisgeneratedatarateof32kb/swithapacketsizeof160bytes.ThesimulationresultsareshowninFig. 5 and 5 ,whereforthroughputeachpointistheaveragedvalueoveronesecondandfordelayeachpointrepresentsonepacket. Fig. 5(a) showsDRAcanprovideaconstantbitrateforCBRvideotrafcwhileFig. 5(a) showsthe802.11failstodoso.Therearenoreal-timepacketdropsobservedinDRAwhileanumberofreal-timepacketsaredroppedduetoMACcollisionsaswellasqueueoverowsin802.11.Thedeliveryratiosofallthereal-timeowsare100%inDRA,andvaryfrom69.3%to97.9%forvideoowsandfrom40.5%to66.8%forvoiceowsinthe802.11.

PAGE 137

(b)Voicedelay (c)Videodelay QoSperformanceinDRA (b)Voicedelay (c)Videodelay QoSperformancein802.11 InDRA,onlyseveralvideopacketshaveadelaylargerthanthe400mslimit.Thereareabout50voicepacketsthathavedelaylargerthan400ms.Ifwelookatthedelayperformanceofeachvoiceow,suchastherstvoiceowwhichstartsat30second,thereareatotalof16packetsthatviolatethedelayrequirement.Theyhappenin4burstseachwith3to5packets.Itmeansthat,duringthetotal600secondsperiod,thereare4shortperiodswhichareabout400ms600mslongeachwith3to5packetsviolatingthedelayrequirement.ThisisacceptableformostofthecurrentVoIPusers.InDRA,themeanandstandarddeviationarerespectively14msand29.8msforvoicepackets,12msand17.9ms

PAGE 138

forvideopackets.99%ofvoicepacketshavedelaybelow115.2ms,anditis74.2msforvideopackets. Therehavebeenmanyresearchworksfocusingonreducingcollisionsandincreas-ingthroughputinWLANs.MACA[ 82 ]andMACAW[ 14 ]usedtheBEBandRTS/CTSmechanismswhichwereadoptedby802.11.Inthepapers[ 16 21 ],anadaptivecontentionwindowwasproposedtoreplacetheBEBmechanism.Bothschemesrelyontheestima-tionofthenumberofactivenodesinthenetwork.Thelatteralsoneedstoestimatethelengthoftransmittedpackets.However,asshowninthepaper[ 17 ],toobtainanaccu-rateandtimelyestimateofthenumberofactivenodesisnoteasy.Therearealsosomeotherschemeswhichdonotneedsuchestimations;however,theysubstantiallychangetheIEEE802.11standard.Forinstance,FCR[ 90 ]usesanewfastcollisionresolutionbackoffschemetoreplacetheBEB.Recently,KimandHou[ 85 ]proposedMFSthatreducesthecollisionsbyschedulingadelaybeforeanodeattemptstransmission.Despitethesigni-cantimprovementover802.11,theachievedthroughputdropssignicantlyasthenumberofnodesincreases.Also,MFSrequiresrun-timeestimationsofthenumberofactivenodes.Inthiswork,withoutchangingtheBEBorinducingtheestimationofthenumberofnodes,weachievehighthroughputbyaccuratelycontrollingthetotaltrafcrateinlightoftheobservedchannelbusynessratio.Morenotably,thethroughputisrathersteadyeveninthepresenceofalargenumberofnodes. Meanwhile,somestudiessoughttoimprovethroughputbyexploitingwirelesschannelvariations[ 80 65 113 76 ].Theunderlyingideaistoincreasetransmissionrateand/ortransmitmorepacketswhenthechannelisgood.However,thisideadoesnotnecessarilyreducethecollisionprobabilityforeachtransmissionattemptsinthepresenceofmanyconcurrentusers.Sotheeffectofcollisiononefciencyisnotnecessarilyalleviated.It

PAGE 139

canbeseenthattheseschemescanbeintegratedintoourDRAframeworktodeliverbetterperformance. SeveralrepresentativeschemeshavestudiedthefairnessissueinWLANs[ 123 110 141 121 8 ].Pilosofetal.[ 110 ]discoveredtheunfavorableeffectofthebuffersizeattheAPondownlinkTCPowsandproposedtoresetthereceiverwindowofalltheTCPowsattheAP.Realizingthatthethroughputfairnessmodelmayleadtoaggregatethroughputdegradationinmulti-rateWLANs,TanandGuttag[ 121 ]proposedtouseatime-basedregulatorattheAPtoachievetime-basedfairnessandhenceimprovethrough-put.However,like[ 110 ],itcanonlyworkininfrastructureWLANs.BymodifyingtheBEB[ 123 ],Vaidyaetal.proposedadistributedfairschedulingschemethatcanimitateSelf-ClockedFairQueueing(SCFQ),acentralizedfairqueueingalgorithm.SCOREwasusedtoachieveproportionaldifferentiation[ 141 ].Again,theBEBprocesswasreplacedbyanadaptiveinter-transmissionspacingcontrol.Moreoverthisschemeand[ 123 ]weredesignedforthroughputfairness.Recently,Bejeranoetal.[ 8 ]studiedthenetworkwideresourceallocationproblemthroughintelligentassociationofuserstoAPs.However,re-sourceallocationamongmobileusersunderoneAPwasnotstudied.Incontrast,DRAisafullydistributedschemethat,withoutchangingtheBEB,achievesthedesirabletimefairness. AlongthelineofQoSsupportinWLANs,mostworksfocusedonservicedifferen-tiation,suchassomerepresentativeschemes[ 1 125 137 ].AdaandCastelluccia[ 1 ]pro-posedtousedifferentinterframespaces,contentionwindowsormaximumframelengthsfordifferentpriorities.Veresetal.[ 125 ]proposedtwomechanisms,i.e.,virtualMACandvirtualsource,toprovidedifferentiatedservices.Toenhancetheemerging802.11estan-dard([ 31 72 ])inQoSsupport,Xiaoetal.[ 137 ]adoptedatwo-levelmechanismtoprotectthereal-timetrafc.SobrinhoandKrishnakumar[ 119 ]proposedtheBlackburstschemetominimizedelayforreal-timetrafc.Unfortunately,stationstransmittingreal-timetrafcarerequiredtohavethechannel-jammingcapability.Inthepaper[ 114 ],thetransmission

PAGE 140

periodissplitforreal-timeandnon-real-timetrafc,therebyenablingQoSguaranteeforreal-timetrafc.However,theDCFmodewasdramaticallychanged.Tosumup,ifthesemanticsofthe802.11DCFismaintained,alltheworksmentionedabovecanonlysup-portservicedifferentiation.Thiswork,however,canprovidestatisticalQoSguaranteeforreal-timetrafcwithoutmodifying802.11. Fromtheabovediscussionsontherelatedwork,weclearlyseethattherehasbeennosystematicstudythataddressesefciency,fairness,andQoSsimultaneouslyon802.11.Tothebestofourknowledge,ourworkistherstonealongthisline.Forthisreason,wemainlyfocusonevaluatingtheperformanceofDRAratherthancompareitwithotherschemesthatonlyaddressedoneoftheseissues.Further,since802.11hasalreadybeenwidelyimplementedincommercialproducts,ourschemethatrunsontopofitismoreattractivethanthosethatdirectlyalter802.11'ssemanticstoalessormoreextent.Here,wehighlightthefollowingadvantagesofDRAoverthoseschemes.Unlikethepreviousschemesthatwereaimedatimprovingefciency,DRAstillyieldshighthroughputeveninthepresenceofalargernumberofusers.Comparedwiththeschemestargetingfairness,DRAachievesgoodtimefairness.Inaddition,DRAgreatlyimprovestheshorttermfair-nessandthemax-minfairness(refertoSection 5.3.2 )especiallywhenthenumberofusersislargerthan10.Insteadofonlysupportingservicedifferentiation,asisthecasewithmostpreviousschemesthatmaintainthe802.11DCFsemantics,DRAprovidesstatisticalQoSguaranteeforreal-timeservice.

PAGE 141

Motivatedbythisobservation,weconductedacomprehensivestudyanddevisedDRA,adistributedresourceallocationschemethatisperhapstherstgeneralframeworkaimedtoaddresshighefciency,timefairness,andQoSsupportforreal-timeservicesatthesametime. DRAutilizesthechannelbusynessratiotocharacterizethenetworkstatus.Basedonthisinformation,DRAdevelopsanovelthree-phasecontrolmechanism,namelythemultiplicative-increase,additive-increaseandmultiplicative-decreasephases,toenablethenetworktoconvergetohighthroughputandtimefairness,whichisprovenbytheoreticalanalysis.QoSforreal-timeservicesisachivedbyconductingcalladmissioncontroloverreal-timeservicesandproperlyadjustingthenetworkresourceamongreal-timeandnon-real-timeservices. ExtensivesimulationsdemonstratethatDRAmaintainshighefciency,goodtimefairness,shortmediumaccessdelayandzeropacketlossrateforreal-timetrafc.Com-paredto802.11,itimprovesthroughputbyashighas71%withRTS/CTSand157%with-outRTS/CTS.Timefairnessisachievedforsingle-rateandmulti-rateWLANs.Moreover,real-timetrafcsuchasVoIPorstreamingvideoissupportedwithstatisticalguarantee.

PAGE 142

Physicalcarriersensingisaneffectivemechanismofmediumaccesscontrol(MAC)protocolstoreducecollisionsinwirelessnetworks,andthesizeofthecarriersensingrangehasagreatimpactonthesystemperformance.PreviousstudieshaveshownthattheMAClayeroverheadplaysanimportantroleindeterminingtheoptimalcarriersensingrange.However,variabletransmissionrangesandreceiversensitivitiesfordifferentchannelratesandtheimpactofmultihopforwardinghavebeenignored.Inthischapter,weinvestigatetheimpactsofthesefactorsaswellasseveralotherimportantfactors,suchasSINR(signaltointerferenceplusnoiseratio),nodetopology,hidden/exposedterminalproblemsandbidirectionalhandshakes,ondeterminingtheoptimumcarriersensingrangetomaximizethethroughputthroughbothanalysisandsimulations.Theresultsshowthatifanyoneofthesefactorsisnotaddressedproperly,thesystemperformancemaysufferasignicantdegradation.Furthermore,consideringbothmultiratecapabilityandcarriersensingranges,weproposetousebandwidthdistanceproductasaroutingmetric,whichimprovesend-to-endthroughputbyupto27%inthesimulatedscenario. 68 ]protocolisakindofCSMA/CA(carriersensemultipleaccesswithcollisionavoidance)MACprotocolsandithasbeenthestandardofthewirelessLANs.The802.11DCF(distributedcoordinationfunction)protocolhasbeenalsowidelystudied 123

PAGE 143

inthewirelessmultihopadhocnetworksduetoitssimpleimplementationanddistributednature. CarriersensingisafundamentalmechanisminCSMA/CAprotocols.Eachusersensesthechannelbeforeatransmissionanddefersthetransmissionifitsensesabusychanneltoreducethecollision.Thismechanismconsistsofphysicalcarriersensingandvirtualcarriersensing.Inthephysicalcarriersensing,thechannelisdeterminedbusyifthesensedsignalpowerislargerthanacarriersensingthresholdCSthoridleotherwise.Inthevirtualcarriersensing,eachuserregardsthechannelbusyduringtheperiodindicatedintheMACheaderoftheMACframes,suchasRTS(readytosend),CTS(cleartosend),DATA,andACK(acknowledgement)denedintheIEEE802.11protocol. Thevirtualcarriersensingmechanismcanonlynotifythenodesinthetransmissionrangeoftheoccupiedmedium,inwhichatransmissioncanbedecodedcorrectlyifthein-terferencelevelissmallenough.Transmissionsoutsideofthisrangecanintroduceenoughinterferencetocorruptthereceptioninmanycases.Inaddition,someongoingtransmis-sionsmaynotbedecodedcorrectlyduetoothertransmissionsnearby,resultinginthefailureofthevirtualcarriersensing.Hencevirtualcarriersensingcannotruleoutcolli-sionsfrominsideofthetransmissionrangeandisincapableofavoidingcollisionsfromoutsideofthetransmissionrange. Physicalcarriersensingrange,inwhichatransmissionisheardbutmaynotbede-codedcorrectly,canbemuchlargerthanthetransmissionrangeandhenceitcanbemoreeffectivethanthevirtualcarriersensinginavoidingtheinterferenceespeciallyinthemul-tihopnetworks.However,largecarriersensingrangereducesspatialreuseandaffectstheaggregatethroughputbecauseanypotentialtransmitters,whichsenseabusychannel,arerequiredtokeepsilent.Therefore,theoptimumcarriersensingrangeshouldbalancethespatialreuseandtheimpactofcollisionsinordertooptimizethesystemperformance. TheIEEE802.11a/b/gprotocolsprovidemultiplechannelratesinwirelessmultihopadhocnetworks.Differentchannelrateshavedifferenttransmissionranges,requirements

PAGE 144

ofSINR(signaltointerferenceplusnoiseratio)andreceiversensitivity.Doeseachraterequireadifferentoptimumcarriersensingthreshold?Howcanwesetthecarriersensingthresholdwhenmultipleratescoexist?Furthermore,multipleforwardingsarecommonformultihopowsandmayforceasignicantchangeoftheoptimumcarriersensingthresh-oldfromthecasewhenonlyone-hopowsareconsidered.Higherchannelratesresultinshortertransmissiondelaybutalsohaveshorterranges.Wemustbecarefultoselecttheappropriatechannelratetomaximizethesystemperformanceintermsofend-to-enddelay/throughputandpowerconsumption,whichareallimportantperformancemetricsformultihopows.Tooptimizetheend-to-endperformanceofmultihopows,carriersensingrangeandspatialreuseaswellashopdistancemustbeappropriatelyaddressed. Thedefaultsettingofthephysicalcarriersensingthresholdandthecarriersensingstrategyinthewidelyusednetworksimulationtoolsns2andOPNETarenotoptimuminmostcases.Theexcessivecollisionsresultinfalselink/routefailuresfollowedbyrerout-ingandunnecessaryend-to-endretransmissionsofTCPpackets.PoorperformanceattheMAClayeraswellasatthehigherlayershasbeenreportedinmanyliteraturesespeciallyformultihopowsinwirelessadhocnetworks([ 91 28 29 151 152 162 147 ]).Fur-thermore,thesesimulationtoolshavenotconsideredthevariablerequirementsofcarriersensingrangesandtransmissionrangeswhenmultiplechannelratesoftheIEEE802.11protocolsareused,hencethesimulationstudiesmaynotreecttheperformanceofrealproducts. Manypapershavealreadynoticedtheimpactofcarriersensingandspatialreuseonthesystemperformance.Xuetal.[ 138 ]indicatethatvirtualcarriersensingviaRTS/CTSisfarfromenoughtosolvetheinterferenceandlargerphysicalcarriersensingrangecanhelpinsomedegree.Inthepapers[ 55 61 54 ],co-channelinterferenceisanalyzedtoderivethespatialreuseandthecapacityofwirelessnetworkswhereinaminimumSINRisnecessaryforsuccessfulcommunication.Gobrieletal.[ 52 ]constructacollisionmodelto-getherwithaninterferencemodelofauniformlydistributednetworktoderivetheoptimum

PAGE 145

transmissionpowerthatyieldsmaximumthroughputandminimumenergyconsumptionpermessage.Lietal.[ 93 ]identifyseveralunfairnessproblemsduetotheEIFSdurationrequiredbythecarriersensingmechanismandproposetousevariableEIFSduration. Recently,severalworkhavealsoattemptedtoidentifytheoptimumcarriersensingrange.Dengetal.[ 37 ]illustratetheimpactofphysicalcarriersensingrangeontheag-gregatethroughputofone-hopowsandproposearewardformulationtocharacterizethetrade-offbetweenthespatialreuseandpacketcollisions.Zhuetal.[ 167 ]haveattemptedtoidentifytheoptimalcarriersensingthresholdthatmaximizesthespatialreuseforareg-ulartopology.YangandVaidya[ 142 ]showthatMAClayeroverheadshaveagreatimpactonthechoiceofcarriersensingrange.However,theinteractionsbetweencarriersensingrangeandvariabletransmissionrangesfordifferentchannelrates,aswellastheirimpactonthenetworkperformance,havenotbeenidentiedbypriorresearch,andtheimpactofmultihopforwardingonthecarriersensingrangehavenotbeenaddressedeither.Therearealsoseveralotherimportantfactorsneededtobefurtherstudiedtodetermineanoptimumcarriersensingrange,suchasvariablerequirementsofSINRandreceiversensitivitiesfordifferentchannelrates,bidirectionalhandshakes,tradeoffbetweenspatialreuseandcolli-sions,nodedensityandnetworktopology,andtheimpactonhigherlayers'performance.Inthischapter,weusebothanalysesandsimulationstoillustratetherelationshipsbetweenallthesefactorsandthesystemperformance.Wedemonstratethatifanyofthesefactorsisnotconsideredproperlyindeterminingtheoptimalcarriersensingrange,thesystemperformancecansufferasignicantloss. Therestofthischapterisorganizedasfollows.Section 6.2 studiestheoptimumcarriersensingrangesubjecttovariousfactorsanditsimpactontheaggregateone-hopthroughput.BasedontheresultsinSection 6.2 ,weillustrateinSection 6.3 howtosetthecarriersensingthresholdinamultirateadhocnetworkandhowitaffectstheend-to-endthroughput,delayandenergyconsumptionofmultihopows.InSection 6.4 ,weintroduce

PAGE 146

severalimportantns2extensionsandconductsimulationstudiestoverifytheanalyticalresults.Finally,Section 6.5 concludesthischapter. Itcanbeshownthatthemaximuminterferencelevelisachievedwhensixothernodesaretransmittingsimultaneouslyattheboundaryofthecarriersensingrangeofeachtrans-mitterasshowninFig. 6 giventhatanytwotransmittersmustbedcawayfromeachother.Similartothecellularnetworksscenario,these6nodesarethersttierinterferencenodes.Sinceanyotherinterferencenodesarefarawayandcontributemuchsmallerinter-ferencethanthersttierinterferencenodes,weignorethemwhencalculatingtheSINR.Tofacilitatethecalculation,wealsoshowthetwo-dimensionalcoordinatesofthenodesinFig. 6 (b),wheredenotestheincludedanglebetweenN0D0andN0N1.

PAGE 147

Figure6: Interferencemodel LetdidenotethedistancebetweennodeNi(06i66)andD0andd06dt.ThenthereceivedpowerPi(06i66)atnodeD0ofthesignalfromnodeNiisequalto whereisthepathlossexponentandtypically2665.SINRisequalto wherePNisthenoiselevelandnormallyitismuchlessthanthepowerleveloftheclosestinterference.Itcanbeshownthatwhenisintherange[0;=6],SINRisanincreasingfunctionofwhen>2.Sinceweconsidertheworstcase,SINRshouldbecalculatedat=0andd0=dt,i.e., (X1)+1 (X+1)+2

PAGE 148

GivenarequirementofSINRforacoding/modulationschemeandthecorrespondingachievablechannelraterc(bps),Xisdetermined.Theachievabledataraterd(bps)equals whereTpreambleinsecondsisapreambleofapacketregardlessofthechannelrate,suchasthephysicallayerpreambleforsynchronizationpurposeatthereceiverandtheshortinterframespacingSIFSattheMAClayer.LHconsistsofprotocoloverheadsinbitsfromdifferentprotocollayers,suchasMACandIPlayers,andLplisthesizeofthepayloadinbitswewishtotransmit. Tocalculatethemaximumaggregatethroughput,weneedtoknowthetotalnumberofconcurrenttransmissions.ForatopologywithanareaofAandwithconcurrentlytrans-mittingpairsasshowninFig. 6 ,eachtransmit-receivepairoccupiesanonoverlappingareaofA0=p 2d2cbyignoringthebordereffect.Forageneraltopology,A0isproportionaltod2c.Thusthetotalnumberofallowedconcurrenttransmissionsis A0/1 ThustheaggregatethroughputSisproportionalto Inthefollowingsubsections,wewilldiscusshowtoselectXtomaximizetheaggregatethroughputusingShannonCapacityandthe802.11datarates,respectively. OnceXisdetermined,CSthcanbesetasthepowerlevelsensedatdistancedctoguaranteeanynewconcurrenttransmissionhappensatleastdcaway.LetTcsdenotetheratioofRXthtoCSth:

PAGE 149

Figure6: CarriersensingthresholdwithShanonCapacity Thus WhenXissmall,log2(1+SINR)increasesalongwithXandisfasterthanX2.WhenXislarge,log2(1+SINR)increasesalongwithXandisslowerthanX2.ThusthereisanoptimumvalueofXtomaximizeS.BylettingthederivationofEquation( 6.11 )withrespecttoXequal0,theoptimumvalueofXcanbesolvedgivenvaluesofTpreamble,LHandLpl.TheresultsareshowninFig. 6 atTpreamble=0andLH=0.

PAGE 150

Table6: Signal-to-noiseratioandreceiversensitivity Rates(Mbps) SINR(dB) Receiversensitivity(dBm) 54 24.56 -65 48 24.05 -66 36 18.80 -70 24 17.04 -74 18 10.79 -77 12 9.03 -79 9 7.78 -81 6 6.02 -82 Figure6: CarriersensingthresholdwithdifferentSINR WhenTpreamble>0andLH>0,XandTcscouldbesmaller.WearenotgoingtofurtherdiscusstheimpactoftheseprotocoloverheadsonthecarriersensingrangeunderShannonCapacity.However,aswewilldiscussbelow,whenthediscretedataratesofthestandardIEEE802.11areconsidered,therequirementofSINR,otherthantheseprotocoloverheads,playsthemajorroletodeterminetheoptimumcarriersensingrange. 6 showstherequirementsofSINRofsomeproductsfordifferentchannelrates[ 143 ].GiventheSINR,wecanderivethevalueofXaccordingtoEquation( 6.4 ).SmallerXviolatestherequirementofSINRandlargerXdecreasesthespatialreuseandtheaggregatethrough-put.Thus,byEquation( 6.7 ),theprotocoloverheadsTpreambleandLHdonotimpacttheoptimumvalueofX.

PAGE 151

Figure6: Carriersensingthresholdwithdiscretechannelratesof802.11 TheoptimumvalueofXandTcsfordifferentSINRaregiveninFig. 6 .WealsoillustrateinFig. 6 theoptimumcarriersensingrangefor4discretechannelrates6,18,36,and54MbpswiththerequirementofSINRinTable 6 .Fromthegure,wecanobservethatthelargertheSINRrequirementis,thelargerXis.XchangesinalargerangefordifferentvaluesofSINRandsodoesTcs. Fortunately,thisdoesnotmeanthattheoptimumcarriersensingrangechangesinalargerangebecausethetransmissionrangeandRXthalsochangeinalargerangefordifferentchannelrates.Table 6 showstherequirementsofreceiversensitivityofthesameproductfordifferentchannelrates.RXthshouldbelargerthanorequaltotherequirementofreceiversensitivity.Thisactuallyrepresentsacommonknowledgethathigherchannelratessustainsinasmallerrangeforthe802.11products.TheoptimumvalueofcarriersensingthresholdCSthcanbeobtainedaccordingtothegivenRXthandthecorrespondingoptimumvalueofTcsforacertainchannelratebythefollowingequation,

PAGE 152

Leti2f6;18;36;54gdenotetheindexofdc(i)andCSth(i)atchannelrate6,18,36,and54Mbps,respectively.Wedeneanormalizedmaximumcarriersensingdistanceastheratioofdc(i)todc(6),representingtherelativesizeoftheoptimumcarriersensingrangeatdifferentchannelratestothatat6Mbps,and BysettingRXthwiththevalueofthereceiversensitivity,weplotCSthanddc(i) 6 .WecanobservethatalthoughtheoptimumvalueofXhasabigdifferenceatdifferentchannelrates,thecarriersensingthresholdandthecarriersensingrangedoesnothavemuchdifferencefordifferentchannelratesandthedifferencerangesfrom0to2dB. 6 .However,thesituationrarelyhappensinpractice.First,inarandomtopology,thepossibilitythatthenodesarelocatedatthedesiredplacesissmall.Second,evenithappens,thechanceisstillsmallforallofthemsuccessfullycontendforthechannelforconcurrenttransmissions.Therefore,considering6concurrentinterferencenodesistooconservativetomaximizethespatialreuseinrandomtopology. Anotherextremeistoonlyconsideronepossiblynearestinterferencenode,likenodeN1tothetransmit-receivepairN0andD0inFig. 6 .Thenearestinterferencedistanceisdcdt,andletX0,CS0thandT0csdenotethecorrespondingX,CSthandTcs,respectively.Wehave GiventherequirementsofSINR,wecanusethesamemethodasthatinSection 6.2.3 toderivethecarriersensingthresholdCS0thfordifferentchannelrates.WehaveX0CSthandT0cs
PAGE 153

Figure6: Tradeoffbetweenexposedterminalproblemandhiddenterminalproblem maximumcarriersensingdistanceis50%to94%,andtheareaofcarriersensingrangeis25%to89%oftheoriginalvaluesdependingonthevaluesofpathlossexponentandchannelraterc.ThusahighercarriersensingthresholdCSthmaygreatlyincreasethespatialreusewithoutintroducingseverecollisions. 6.2.2 and 6.2.3 ,wehaveconsideredalltheinterferenceattheworstcasetocalculateXandCSth.Therewillbealmostnohiddenterminalprob-lems,becauseallinterferencenodeswhichmaycontributeenoughinterferencetocorruptthereceivedpacketsfallinthecarriersensingrangeandarerequiredtodefertheirowntransmissions.Noticethatthephysicalcarriersensingdoesnotcompletelysolvethehid-denterminalproblem.ForexampleinFig. 6 (b),thereisanobstructbetweennodeAandC.CmaynotbeabletosensethetransmissionofAandhencemayinitiateanewtransmission,resultinginacollisionatnodeB. Aswehaveindicatedintheprevioussubsectionthatthiscarriersensingthresholdmaybetooconservative.Here,wepointoutanotherreasonthatthelargecarriersensingrangeimpactsthespatialreuse.AsshowninFig. 6 (a),nodeC'stransmissiontoDwillnotintroduceenoughinterferencetocorruptB'sreception.However,nodeCsenses

PAGE 154

A'stransmissionanddefersitsowntransmission.Thisresultsinpoorspatialreuseandiscommonlycalledtheexposedterminalproblem. FromthepointofviewoftheintendedreceiverBinFig. 6 (a),wecalltherangeastheinterferencerangeofB,inwhichanypointhasacloserdistancetoBthan(X1)dt.AnysingleinterferencenodeoutofthisrangemaynotcorruptB'sreceivedpacketfromA.Tomeasuretheimpactoftheexposedterminalproblem,wedeneanexposed-arearatioastheratiooftheareaofcarriersensingrangetothatofinterferencerangeminus1and X121(6.15) Itiseasytoshowthatdecreasesfrom3to0.05whenXincreasesfrom2to40.FromFig. 6 ,weknowthatasmallerchannelraterequiresasmallerXandhencealargerexposed-arearatio.Andevenforthehighestchannelrate54Mbpsallowedin802.11a/g,theexposed-arearatiocannotbeignoredwhen>3because=24%and56%whenX=10and5,respectively.Letalonethatwehavenotconsideredaworsecasebutacommonsituationinarandomtopologywherethedistancebetweenthetransmitteranditsintendedreceiverislessthanthemaximumtransmissiondistancedt.Theinterferencerangeshouldbesmallerbecausethereceivedsignalhasagreaterpower.Therefore,theexposedterminalproblemcannotbeignored. Therefore,toalleviatetheexposedterminalproblemandincreasethespatialreuse,itisnecessarytodecreasethecarriersensingrange.However,thismayexposeapartoftheinterferencerangeoutofthecarriersensingrangeandresultsinhiddenterminalproblem.Thesmallerthecarriersensingrange,thelesstheexposedterminalproblemandthemoreseverethehiddenterminalproblem.Apparently,thereisatradeoffbetweentheexposedterminalproblemandthehiddenterminalprobleminordertoincreasespatialreuseandalleviatethecollisionsatthesametime.

PAGE 155

6.2 focusesonone-wayDATAtransmis-sions.However,wirelesslinksarenotreliableduetocollisions,wirelesschannelerrorsandmobility.MAClayeracknowledgementsarenecessarytocheckthelinkreliabilityandinmostcases,linklayerretransmissionsaremoreefcientthanend-to-endretransmissionsforsuchunreliablelinks.Besides,thebidirectionalhandshakehasalreadybeenadoptedbytheIEEE802.11MACprotocolswhichdeneatwo-wayDATA/ACKhandshakeandafour-wayRTS/CTS/DATA/ACKhandshakeforeachdatapackettransmission.There-fore,thereceiversmayalsotransmitCTS/ACKandtheinterferenceatnodeD0inFig. 6 becomesworse,hencetheoriginalinterferencemodelinEquation( 6.3 )isnotalwayseffectivetoavoidthecollisions. Whenbidirectionalhandshakesareconsidered,thefollowingthreeproblemshavenotbeenwelladdressedinthatinterferencemodel.Therstproblemisthepacketcollision.Thesecondoneisthereceiverblockingproblem[ 165 163 ].Thisproblemdescribesthesit-uationthatthetransmitterkeeps(re)transmittingRTSorDATAframeswhentheintendedreceiversensesabusychannelduetootherongoingtransmissionsanddoesnotorcan-notrespondwithCTSorACKframes.Aftertheretransmissiontimesexceedsacertainthreshold[ 68 ],thetransmitterwilldropthedatapacket,declarethelinkfailureandhencerouterepairwillbeexecuted.Wecallthisreceiverasablockedreceiver.Thethirdoneisunfairnessresultingfromtheprevioustwoproblems. Theseproblemsarerelatedwiththecarriersensingstrategies.TherearelargelytwocarriersensingstrategiesintheIEEE802.11.ThestrategyIistoforbidanodefromtransmittingifitsensesabusychannel.ThestrategyIIistoallowanodetotransmitatanysituationsevenifitsensesabusychannel.Inthe802.11,therststrategyisadoptedfortransmissionsofRTS,DATAandCTSframes.ThesecondoneisadoptedfortransmissionsofACKframestoacknowledgethesuccessfullytransmittedDATAframes.

PAGE 156

Weuseasimpletopologytoillustratetheseproblems,i.e.,thelineartopologyABCD.Werstdiscussthepacketcollisionproblem.SupposeDcannotsenseA'stransmissionandisoutofB'sinterferencerange.However,CiscloseenoughtoBsothatC(orB)'stransmissionwillcorruptB(orC)'sreception.WhenAistransmittingtoB,DmayinitiateatransmissiontoC.AnditisthesameforAwhenDistransmittingtoC.IfoneorbothofAandDaretransmittingDATAframes,BorC,whichrstnishesthereceptionoftheDATAframe,willreturnanACKframewhichdoesnotrequirecarriersensingbeforehandandwillcorruptthereceptionattheother.Toalleviatethisproblem,wemayproposetousetheshortframesRTS/CTSbeforeDATA/ACKsinceCTSrequirescarriersensingbeforehand.However,thecarriersensingstrategyofCTSframesmakesthereceiverblockingproblemworse. ThereceiverblockingproblemexistswhentheintendedreceiverdoesnotreturnanACKframeduetoacollision.ItalsoexistsforthecarriersensingstrategyIevenwhentheintendedreceivercancorrectlydecodethereceivedpacketbutcouldnotrespondduetothecarriersenserule.Forexampleintheprevioustopology,supposethatAandCareoutofeachother'ssensingrange,andBandCareoutofeachother'sinterferencerangeandcannotcorrupteachother'sreceivedpackets.However,BandCcansenseeachother'stransmission.WhenCistransmittingalongDATAframetoD,AmayinitiateatransmissiontoBbyaRTSframe.SinceBsensesabusychannel,itdoesnotreturnaCTSframesothatAkeepsretransmittingtheRTSframe.AalsodoublesitscontentionwindowsizeforeachfailureofRTStransmissionandhaslowerchannelaccessprobability.IfChasalotofdatapacketsdestinedtoDandoccupiesthechannelforalongtime,Awillbestarved.ApparentlythisisunfairtoAduetotheMACcontention. Toalleviatetheseproblemsasmuchaspossibleforthebidirectionalhandshakes,itisnecessarytobemoreconservativetosetthecarriersensingrangethantheinterferencemodelinFig. 6 especiallyfortheinterferencemodelattheworstcase.First,toaddressthepacketcollisionproblem,weneedtoconsidertheinterferencesfromnodeDi(16

PAGE 157

6.2.1 ,theSINRattheworstcasesatises bX bX ThenumericalresultsshowthatbXcanbewellapproximatedbyX+1withlessthan1%errorwhenSINRislargerthan-3dB: Forexampleinthepreviousfour-nodetopology,thetworeceiversBandCarethusfarenoughfromeachotherandtheACKframescannotcorrupteachother'sreception.Second,whenRTS/CTSareused,usingthecarriersensingstrategyIIforCTS/DATA/ACKframestoaddressthereceiverblockingproblemandRTSstilladoptsthecarriersensingstrategyI.Forexampleinthesametopologyasbefore,BisfarenoughfromthetransmittingnodeCandcancorrectlydecodeaRTSorDATAframefromA.Moreover,itdoesnotpreventfromreturningaCTSorACKframe.ThenewcarriersensingrangeisshowninFig. 6 Apparently,thecostistoaggravatetheexposedterminalproblemandsacricethespatialreuseinamoregeneraltopology.However,thepacketcollisionanddroppingduetohiddenterminalsandblockedreceivershavebeensignicantlyimproved.Moreoverweexpectamorestableperformanceforhigherlayerprotocols,suchasmuchlessretrans-missionandtimeoutsforTCPtrafc,andmuchlessfalselinkfailuresandunnecessaryreroutingactivities,andtheunfairnessduetothesetwoproblemscanalsobegreatlyalle-viated. Otherproblemsforthelargecarriersensingrangeareasfollows.First,itrequiresasmallsensingthreshold.Wedonotknowtheachievablecarriersensingsensitivityofthecurrentproducts.Inthefollowingstudies,weassumethatthecurrentproductsor

PAGE 158

Figure6: LargecarriersensingrangewithcarriersensingstrategyIIforCTS/ACK futuretechnologiescouldsupportthesmallcarriersensingthreshold,i.e.,Tlcstimesmoresensitivethantheoriginalvalue.ThenewvalueofTcsiscTcsandequals X)(6.18) When=3,Tlcs=5:28;3:75and2:91dBforX=2;3and4,respectively.Second,thelargercarriersensingrangemeansthattheremayexistmorenodescontendingforthesharedchannel.ThecollisionprobabilitylikethatinwirelessLANsincreaseswiththenumberofactivenodes[ 160 154 ]. Toaddressthecollisionprobleminonecarriersensingrange,therearealreadysev-eralmethods.First,four-wayhandshakeinsteadoftwo-wayhandshakecouldbeusedtoreducethelongcollisionperiodsofDATAframetransmissionsifthecollisionprobabilityishighandtheDATAframeislong.Second,someschemes[ 149 148 150 27 26 164 ]controllingthetrafcdeliveredtotheMAClayeraccordingtothechannelstatuscanbeusedtoefcientlyreducethecollisionprobability.Third,wecanmaintainthevalueofXbutreduceboththecarriersensingrangeandtransmissionrangeinordertoreducethenodedensityineachcarriersensingrange.

PAGE 159

SimulationstudieswithconsiderationsofallpreviouslyaforementionedfactorswillbeconductedinSection 6.4 toidentifytheoptimumvalueofcarriersensingthreshold.

PAGE 160

Figure6: Multiplecarriersensingthresholdsmayresultincollisions issubjecttodistance,mobilityandchannelfadingandistimevariable.Hencemultiplecarriersensingthresholdsfordifferentchannelratesmaygreatlyincreasethecomplex-ityoftheprotocols.Second,asdiscussedabove,theoptimumcarriersensingthresholdsdonotchangemuchfordifferentchannelrates.Asinglethresholdwillnotsacricetheperformancemuch.Third,multiplecarriersensingthresholdsmayintroduceadditionalcollisions.ForexampleinFig. 6 ,atransmit-receivepairAandB,whichhavealargecarriersensingrangecorrespondingtoacertainchannelrate,sensesanidlechannelandthenAtransmitstheDATAframes.Duringthetransmissionperiod,anothertransmitterCintheprevioustransmitter'ssensingrangealsosensesanidlechannelduetoasmallercar-riersensingrange.ThenewtransmissionfromCmayintroduceacollisionatthepreviousintendedreceiverB. Withacommoncarriersensingthreshold,thereceivethresholdRXthmustbesetappropriately.First,RXthmustbelargerthanorequaltothereceiversensitivityRXserequiredbytheadoptedchannelrate.Second,toalleviatecollisionsasmuchaspossible,onemorerequirementmaybeenforced,i.e.,thepowerlevelofthereceivedsignalmustbelargerthanorequaltoCSthTcs.Accordingtothesetworequirements,wecansetthecommoncarriersensingthresholdCSthas

PAGE 161

whereiistheindexofdifferentchannelrates. Inwirelessmultihopadhocnetworks,destinationsareoftenoutofthesources'trans-missionrangeandpacketsneedtobeforwardedthroughmultiplehopsbeforereachingthedestinations.Selectingthenexthopwiththehighestchannelratecanincreasethethroughputateachhop.However,packetsmusttravelthroughmorehopsduetotheshorttransmissionrangeofhighchannelratesandhencetheend-to-enddelayandthroughputarenotnecessarilyimproved.Todeterminethebestcandidateofthenexthop,itisneces-sarytointroduceametricconsistingofinformationofbothchannelratesandhopdistances.

PAGE 162

whereiistheindexofhopsalongthepath.Andthehoptransmissiondelaythis SupposeRTS/CTS/ACKaretransmittedwiththebasicrateandDATAistransmittedwiththeselectedchannelraterc,then Todeterminetheefciencyofeachhop(orthatofthecandidatesateachhop),wedeneabandwidthdistanceproductBDiPforeachhopastheachievabledataraterdtimesthehopdistancedhatthathop,then Thepermetertransmissiondelaytmforthehopisequalto End-to-enddelayisthesummationoftransmissiondelayatallforwardingnodes.Ifthepathisaregularchainwhereeachhophasthesamedistanceandthetotalpathlengthisdp,thentheend-to-endtransmissiondelayte2eisequalto Normally,dpisproportionaltothedistancedsdbetweenthesourceandthedestination.Supposeithasarelativelyxedvalue,thentheend-to-enddelayisinverselyproportionaltothebandwidthdistanceproductBDiP.

PAGE 163

Figure6: Bandwidthdistanceproduct Inthischapter,weassumeacommontransmissionpowerPtforallchannelrates.ThustheaggregatetransmissionpowerconsumptionEforeachpacketis SinceTSIFSandTDIFSaremuchsmallerthanTDATA,minimizingtheend-to-enddelayisalmostequivalenttominimizingtheend-to-endenergyconsumptionE. Therefore,tominimizetheend-to-enddelayandenergyconsumption,weshouldse-lectthecandidatewiththehighestvalueofBDiPasthenexthopifotherconditionsarethesame.Fig. 6 showsthebandwidthdistanceproductforseveralchannelrates.Toplotthegure,theadvertisedtransmissionrangesforoutdoorenvironmentsofoneCiscoproduct[ 32 ]areused.Theyare76,183,304,396,and610mfor54,18,11,6,and1Mbpsrespectively.Noticethat1and11Mbpsarethe802.11bratesand6,18and54Mbpsarethe802.11grates.Weusethedefaultparametersdenedforthe802.11b/gratesaccordingtothecorrespondingstandards[ 69 ]and[ 70 ],respectively(802.11grateshaveshorterpream-bles).Twocasesareconsideredwithandwithoutprotocoloverheads.Forthecasewithoutprotocoloverheads,rd=rc.Forthecasewithprotocoloverheads,two-wayhandshakeDATA/ACKand1000bytespayloadsizeareused.

PAGE 164

Figure6: Maximumend-to-endthroughputfordifferenthopdistance TwoimportantobservationscanbefoundfromFig. 6 .First,largerprotocolover-headsresultinsmallerBDiP.Second,higherchannelratesdonotnecessarilygeneratelargerBDiP.ThemaximumvalueofBDiPiscloselyrelatedtobothhopdistanceandpro-tocoloverheadsinadditiontotheachievablechannelrate. Tomaximizetheend-to-endthroughputofamultihopow,itisnecessarytomaxi-mizethespatialreusealongthepath,i.e.,toscheduleasmanyconcurrenttransmissionsatdifferenthopsaspossible.Therearetworequirementstoschedulesuccessfulconcur-renttransmissions.First,twoneighboringtransmittersalongthepathmustbeatleastdcawayfromeachothersothatthereisonlyonetransmitterineachcarriersensingrangetosatisfythecarriersensingrequirement.Second,theconcurrenttransmissionsatupstreamanddownstreamnodescannotintroduceenoughinterferencetocorruptthereceptionattheconsideredtransmit-receivepair.

PAGE 165

Let1 wheredxeistheceilingfunctionofxandisequaltothenearestintegergreaterthanorequaltox.Thus,forachaintopologywithacommondistancedhateachhop,themaximumend-to-endthroughputforamultihopowwithatleastNhopsis becausetherecanbeonlyonesuccessfultransmissionineachNhopsandhencethespatialreuseratioforthechaintopologyis1 Theequalityintheabovetwoinequalitiesholdsonlywhenthecarriersensingrangeissettosatisfythesecondrequirementdiscussedabove.Thatistosay,thereshouldbenohiddenterminalproblemorreceiverblockingproblemasdiscussedinSection 6.2.6 duetothetransmissionatNhopsawayalongthepath.Thismaximumend-to-endthroughputisshowninFig. 6 ,wherewesupposedc=1400mistheminimumvaluetosatisfytheaboverequirements.Whenprotocoloverheadsareconsideredand1000bytespayloadisused,Smaxequalsto1.68,1.79,1.33,1.34,0.30Mbpsfor54,18,11,6,1Mbpsattheircorrespondingmaximumhopdistance,respectively.Itveriesthathigherchannelratesdonotnecessarilygeneratehigherend-to-endthroughput.Itiscloselyrelatedtotheachievablechannelrate,hopdistance,carriersensingrangeandprotocoloverheads. AsdiscussedinSection 6.2 ,theoptimumcarriersensingrangemayallowacertainlevelofhiddenterminalproblemtobalancetheimpactofexposedterminalproblem.Inthiscase,Equations( 6.28 )and( 6.29 )onlyprovidealowerboundforNandanupperboundforthemaximumend-to-endthroughput.Toaccuratelycalculatethemaximumend-to-endthroughput,Nshouldberecalculatedwiththerequirementthattheconcurrentscheduledtransmissionsshouldnotintroduceenoughinterferencetoeachothertocorrupt

PAGE 166

Figure6: Spatialreuseratioformultihopows(a)atworstcase,(b)inasinglechaintopologywithonewaytrafc thereceptions.Thus,NisdeterminedbytherequirementofSINRandthelocationsofthesourcesandforwardingnodes.HereweusetheinterferencemodelusedinEquation( 6.16 )fortheworstcasewithbidirectionalhandshakes,and Ifallhopdistancesequalthesamevalueorthehopdistancesaresetasdh,thesetwoequationscanbesimpliedasN6dbXeandSmax>rd Generally,therearelessinterferencesthantheworstcase.Foraregularchaintopol-ogywithacommonhopdistance,ifweonlyconsidertheinterferencefromonenearestupstreamtransmissionandonenearestdownstreamtransmission,andletcX0denotethevalueofbX,Equation( 6.16 )becomes

PAGE 167

whereSINRisworseforthecaseoftwo-waytrafcbecausethereceiveroftheconcur-rentdownstreamtransmissioncanbecloserthanitsintendedtransmittertotheconsideredreceiver.Thustheaggregateend-to-endthroughputfortwo-waytrafccanbelowerthanone-waytrafc.However,ifanoptimumpacketschedulerispossibletoschedulefor-wardingtrafcatonetimeandreversetrafcatanothertime,theaggregateend-to-endthroughputoftwo-waytrafccanbeashighasthatofonewaytrafc.Thusonlythecaseofone-waytrafcisdiscussedthereafter. SinceasmallervalueofXthancX0dosenothelpincreasethethroughputduetotherequirementofSINRandonlyresultsincollisionsduetothehiddenterminalproblem.Therefore,cX0istheoptimumvalueofXforamultihopowinaregularchaintopologyand dhe6rd Thereforetheachievablemaximumend-to-endthroughputofamultihopowisrd 6.2 Fig. 6 showsbothbXandcX0alongwithdifferentrequirementofSINR.WhenSINR=10dBand=4(fordistanteld)whicharethedefaultsettingsinns2,thespatialreuseratiois1 3andhencethemaximumend-to-endthroughputis1 3ofthebandwidthforachaintopologywithatleast3hops.Thisislargerthanthendingsinthepapers[ 91 162 ]whichshowsthespatialreuseratiois1 4.Therearetworeasonsforthethroughputloss.First,thesepapersstudythefour-wayhandshakeRTS/CTS/DATA/ACKandthethroughputsuffersfromthereceiverblockingproblemaswediscussedinSection 6.2.6 .Second,thesepapersusens2forsimulationstudiesandaMACframeisdiscardedifthereisalreadyaninterferencewhenreceivingtherstbitoftheframeevenwhenSINRishighenoughinthecurrentversionofns2.Fig. 6 alsoshowsthatlargercanachievebetterspatialreuseratioandhencehigherend-to-endthroughputbecausetheinterferencevanishesmore

PAGE 168

quicklyalongwiththedistance.However,largerresultsinshortertransmissiondistance,andhencerequiresmoreforwardingsandconsumesmoreenergyforeachpackettoreachthedestination. Furthermore,SmaxintheaboveequationsonlyconsidermultihopowswithatleastNhops.Foramultihopowwithfewerhops,wehave wherenhisthenumberofhopsofthemultihopow. Inshort,tomaximizetheend-to-endthroughputofamultihopow,itisnecessarytoselectanodewiththehighestvalueofrd 6.33 ),wecanseethatSmaxisap-proximatelyproportionallytothebandwidthdistanceproductBDiP=rddh.ThusBDiPcanbeutilizedtoapproximatelyrepresenttheefciencyofthroughput,delayandenergyconsumptionateachhop.Wewillevaluatetheefciencyofthismetrictomaximizetheend-to-endthroughputandcompareitwiththeshortesthopalgorithmforamultiratenet-workthroughsimulationsinnextsection.

PAGE 169

6.2.6 )inthefollowingsubsections. WeadopttherequirementsofSINRandreceiversensitivitiesinTable 6 unlessotherwiseindicated,andthereceiverthresholdissetasthevalueofthecorrespondingreceiversensitivityforeachchannelrate.Thedefaulttworaygroundpropagationmodelinns2isused,i.e.,thepathlossexponent=2whenthedistanceislessthan86mand=4otherwise,andthetransmitpowerissetas6dBm.Thetransmissionrangesarehencedetermined.Inthesimulations,thechannelrates54,36,18,and6Mbpsarestudied,andtheirtransmissionradiiare89,119,178,and238m,respectively.TheIEEE802.11a[ 71 ]protocolparametersareadoptedinthesimulations. FirstweidentifytheoptimumcarriersensingthresholdCSthforone-hopows.Inthesimulation,eachnoderandomlyselectsoneneighborasthedestinationofoneTCPconnection.Noticethattheneighborhoodissmallerforahigherchannelrateduetoits

PAGE 170

smallertransmissionrange.Fig. 6 showsthattheaggregatethroughputachievesthemaximumvaluewhenCSthisintherangeof[61;76]dBmforallchannelrates.However,suchCSthisevenlessthanRXthforseveralchannelrates.Apparently,itstarvestheowswhosesourcedestinationdistanceisclosetothetransmissionradiusasfoundfromthemoredetailedsimulationresults. WealsoidentifytheoptimumcarriersensingthresholdCSthifmultihopowsexist.Inthesimulation,therearetotal20TCPconnections.Thesourcesandthedestinationsarerandomlyselectedundertheconditionthatthedistancebetweenthesourceandthedestinationrangesfrom500to600m.Thedistanceconditionisusedinsteadofthehopnumberbecausewealsowanttochecktheefciencyofdifferentchannelratestodelivertrafcoverthesamedistanceandhigherchannelratesoftentravelmorehopstoreachthesamedestination.Fig. 6 showsthattheaggregateend-to-endthroughputachievesthemaximumvaluewhenCSthisaround91dBmforallchannelrates.WhentheCSthislessthanthereceiverthresholdRXth,theend-to-endthroughputisalmostzero.Thisisbecausethatsomehopdistancesapproachthemaximumtransmissiondistance,whichleadstodisconnectionsofthesehops. Thereareseveralimportantobservationsfromtheseresults.First,todeterminetheoptimumcarriersensingrangeinthemultihopadhocnetworks,itisnotenoughtoexaminetheperformanceofone-hopowsandtheimpactofmultihopforwardingmustbecarefullystudied.Second,asinglecarriersensingthresholdcouldbeoptimumforallchannelrates.Third,ahigherchannelratedoesnotnecessarilygenerateahigherthroughput.Wemustbecarefultoutilizethemultiratecapabilityinthemultihopenvironmentwhichwillbefurtherstudiedinnextsubsection.Theseobservationsverifyourearlieranalyticalresultsinthischapter.

PAGE 171

Figure6: Optimumcarriersensingthresholdforone-hopows Figure6: Optimumcarriersensingthresholdformulti-hopows

PAGE 172

3insteadof1 4usingthedefaultparametersofns2wheretheSINRre-quirementis10dB.Thehopdistanceissetasthemaximumtransmissiondistanceandthechanneldatarateis6Mbps.Themaximumthroughputisfoundbygraduallyincreas-ingthecarriersensingthresholdandtherateofCBRtrafcfromthesource.Aslongas12dB
PAGE 173

achievableratebetweenthisnodeandthetransmitter.Thesecondoneselectsthefarthestnodeamongthosewithasamehighestchannelrateastheforwardingnode.ThethirdoneselectsthenodewiththehighestvalueofBDiPasthedownstreamforwardingnodeateachhop.TheyarereferredasAdr(rstconsiderthedistance,thentherate),Ard(rstcon-sidertherate,thenthedistance)andABDiP(maximizethebandwidthdistanceproduct),respectively,inthefollowingdiscussions. Themaximumend-to-endthroughputsareachievedwhenPt

PAGE 174

3insteadof1 4forthe10dBSINRrequirement.

PAGE 175

IEEE802.11MACprotocolhasbeenthestandardforWirelessLANsandisalsoimplementedinmanysimulationsoftwareformobileadhocnetworks.However,IEEE802.11MAChasbeenshowntobequiteinefcientinthemultihopmobileenvironments.Besidesthewell-knownhiddenterminalproblemandtheexposedterminalproblem,therealsoexiststhereceiverblockingproblemwhichmayresultinlink/routingfailuresandunfairnessamongmultipleows.Moreover,thecontentionandinterferencefromtheup-streamanddownstreamnodesseriouslydecreasethepacketdeliveryratioofmultihopows.Alltheseproblemscouldleadtotheexplosionofcontrolpacketsandpoorthroughputperformance.Inthischapter,werstanalyzetheseanomalyphenomenainmultihopmobileadhocnetworks.Then,wepresentanoveleffectiverandommediumaccesscontrol(MAC)protocolbasedonIEEE802.11MACprotocol.ThenewMACprotocolusesanout-of-bandbusytoneandtwocommunicationchannels,oneforcontrolframesandtheotherfordataframes.Thenewlydesignedmessageexchangesequencepro-videsacomprehensivesolutiontoalltheaforementionedproblems.Extendedsimulationsdemonstratethatourschemeprovidesamuchmorestablelinklayer,greatlyimprovesthespatialreuse,andworkswellinreducingthepacketcollisions.Itimprovesthethroughputby8%to28%forone-hopowsandby25timesformultihopowsunderheavytrafccomparingtotheIEEE802.11MAC. 68 ]issuchaprotocolthathasbeenthestandardofwirelessLANsandhasalsobeenin-corporatedinmanywirelesstestbedsandsimulationpackagesformobileadhocnetworks. 156

PAGE 176

Itadoptsfour-wayhandshakeprocedures,i.e.,RTS/CTS/DATA/ACK.Shortpackets,RTSandCTS,areusedtoavoidcollisionsbetweenlongdatapackets.TheNAV(NetworkAllocationVector)valuecarriedbyRTS/CTS/DATA/ACKisusedtoreservethemediumtoavoidpotentialcollisions(i.e.,virtualcarriersensing)andhencemitigatethehiddenter-minalproblem.TheACKisusedasaconrmationofthesuccessfultransmissionwithouterrors. However,theeffectivenessofIEEE802.11MACinmultihopmobileadhocnetworkshasbeenwidelyrecognizedasaseriousproblem.ThepacketcollisionovertheairismuchmoresevereinthemultihopenvironmentsthanthatinthewirelessLANs[ 68 150 160 161 153 ].ThepacketlossesduetosuchkindofMAClayercontentionswilldenitelyaffecttheperformanceofthehighlayernetworkingschemessuchastheTCPcongestioncontrolandroutingmaintenancebecauseanodedoesnotknowwhetheranerrorisduetothecollisionortheunreachableaddress[ 19 108 140 161 153 28 29 147 162 ]. ThesourceoftheaboveproblemscomesmainlyfromtheMAClayer.Thehiddenterminalsintroducecollisionsandtheexposedterminalsleadtolowspatialreuseratio.Besidesthesetwonotoriousproblems,thereceiverblockingproblem,i.e.,theintendedreceiverdoesnotrespondtoRTSorDATAduetotheinterferenceorvirtualcarriersens-ingoperationalrequirementsfromotherongoingtransmissions,alsodeservesaseriousattention.Thisproblembecomesmoresevereinthemultihopenvironmentsandresultsinpacketdropping,starvationofsometrafcowsornodes,andnetworklayerre-routing,whichwewillelaboratelaterinsection 7.3 .Furthermore,formultihopows,thecon-tentionsorinterferencesfromtheupstreamanddownstreamnodesandotherowscouldleadtopoorpacketdeliveryperformance. Therearemanyschemesproposedinthecurrentliteraturetoreducetheseverecolli-sionsofDATApacketsatMAClayer.BTMA[ 122 ]usesabusytonetoaddressthehiddenterminalproblem.Thebasestationbroadcastsabusytonesignaltokeepthehiddenter-minalsfromaccessingthechannelwhenitsensesatransmission.Itreliesonacentralized

PAGE 177

networkinfrastructurewhichisnotapplicableinmobileadhocnetworks.FAMA-NCS[ 47 ]usesthelongdominatingCTSpacketstoactasthereceivebusytonetopreventanycompetingtransmittersinthereceiverrangefromtransmitting.Thisrequiresanynodeshearinginterferencekeepquietfortheperiodofonemaximumdatapackettoguaranteenocollisionswiththeongoingdatatransmission,whichisobviouslynotefcientespeciallywhentheRTS/CTSnegotiationprocessfailsortheDATApacketisveryshort. Somemulti-channelschemesbasedonrandomaccesshavealsobeeninvestigatedinthelastfewyears.Onecommonapproachtoavoidthecollisionsbetweencontrolpacketsanddatapacketsistouseseparatechannelsfordifferentkindsofpackets.DCA[ 136 ]usesonecontrolchannelforRTS/CTSandoneormoredatachannelsforDATA/ACK.Itpresentsonemethodtoutilizemultiplechannelsbutdoesnotsolvethehiddenterminalproblems.Dualbusytonemultipleaccess(DBTMA)schemes[ 59 60 135 ]handlesthehiddenterminalandexposedterminalproblems.Itusesthetransmitbusytonetopreventtheexposedterminalsfrombecomingnewreceivers,thereceivebusytonetopreventthehiddenterminalsfrombecomingnewtransmitters,andaseparatedatachanneltoavoidcollisionsbetweencontrolpacketsanddatapackets.DBTMA,however,doesnotconsidertheACKpacketswhich,ifused,mayresultincollisionswiththeDATApacketswhiletheacknowledgment(ACK)isneededfortheunreliablewirelesslinks.PAMAS[ 117 ]usesaseparatecontrolchanneltotransmitbothRTS/CTSpacketsandbusytonesignals.Itgivesasolutiontothehiddenterminalproblemandmainlyfocusesonpowersavings.MAC-SCC[ 92 ]usestwoNetworkAllocationVectors(NAVs)forthedatachannelandthecontrolchannel,respectively.ThetwoNAVsmakeitpossibleforthecontrolchanneltoschedulenotonlythecurrentdatatransmissionbutalsothenextdatatransmission.Althoughitreducesthebackofftime,itdoesnotaddresstheaforementionedproblems. Tothebestofourknowledge,therearenocomprehensivestudyandgoodsolutionstoallthehiddenterminalproblem,theexposedterminalproblem,thereceiverblockingproblem,andtheintra-owandinter-owcontentionproblems.Allofthemcontributeto

PAGE 178

thepoorperformanceofMACprotocolinthemultihopwirelessmobileadhocnetworks.Mostofthecurrentschemesaggravatethereceiverblockingproblemwhilealleviatingthehiddenterminalproblemanddonotfullyaddresstheproblemsofmultihopowsinthemobileadhocnetworks. Inthischapter,weutilizetwochannels(dual-channel)forcontrolpacketsandDATApackets,separately.RTSandCTSaretransmittedinaseparatecontrolchanneltoavoidthecollisionswithdatapackets.NegativeCTS(NCTS)isusedtosolvethereceiverblockingproblemandisalsotransmittedinthecontrolchannel.Anoutbandreceiver-basedbusytone[ 59 ]isusedtosolvethehiddenterminalproblem.WedonotuseACKherebecausethereisnocollisiontotheongoingDATApacket.Toaddressthepacketerrorduetotheimperfectwirelesschannel,weintroduceNegativeAcknowledgment(NACK)signal,acontinuingbusytonesignal,whenthereceiverdeterminesthatthereceivedDATApacketiscorruptedandinerror.ThesenderwillnotmisinterpretthisNACKsignalbecausetherearenootherreceiversinitssensingrangeandhencenointerferingNACKsignals,anditwillassumethatthetransmissionissuccessfulifnoNACKsignalissensed.Furthermore,ourprotocolhasaninherentmechanismtosolvetheintra-owcontentionandcouldachieveoptimumpacketschedulingforchaintopology.ItturnsoutthatthisprotocolhassolvedalmostallaforementionedproblemsanddoesnotrequiresynchronizedtransmissionattheMAClayerasinthepapers[ 7 118 ]. Therestofthischapterisorganizedasfollows.Section 7.2 presentsthebasicconceptsofthephysicalmodelwhichareimportanttodesigntheMACprotocol.Then,Section 7.3 elaboratesthesourceofcollisionsintheIEEE802.11MACprotocolwhenappliedinthemulti-hopmobileadhocnetworksandtheidealprotocolbehaviorwemaydesire.Section 7.4 describesthenewMACprotocolformultihopmobileadhocnetworks.SimulationresultsaregiveninSection 7.5 .Finally,weconcludethechapterinsection 7.6

PAGE 179

7.2.1PhysicalModel ThereceivedpowerPr: wheredoisthereferencedistanceandPoisthereceivedpoweratthereferencedistance.>2isthepower-lossexponent.Inthefollowingdiscussions,weassumeallnodesusethesametransmissionpower. Ifthereisinterferencefromanothertransmissionatthereceiver,thepoweroftheinterferencesignalPimustbesmallerenoughthanthatoftheintendedsignalPr,i.e.PiCPThresh1isthecapturethreshold.So wherediisthedistancefromtheinterferencesourcetothereceiver,anddristhedistancefromthesendertothereceiver.Thequantityc=CP1=Thresh>1denesazonewhereothertransmissionswillinterferethereceivingactivities.

PAGE 180

Whenthereceiverismaximumtransmissiondistancedtawayfromthesender,theminimuminterferencedistance,dimin,whichallowcorrectdemodulationatthereceiverandtheinterferencepowerPiminare SothesendershouldbeabletosensetheinterferencewithpowerlevelPiminbeforetransmission,i.e.,theinterferencefromdiminaway,toavoidpotentialinterferencetootherongoingtransmission.Consideringtheprobabilitythattherearemorethanoneinterferingtransmissionsintheneighborhoodoftheintendedreceiver,thesensingrangedsshouldbeevenlargerthandimin,i.e., whichcanguaranteecorrectreceptionatthereceiverifitsensesthechannelidleinspiteofthepossibleinterferencesfrommultiplesourcesoutsideofthesensingrange. Thesensingrangeisalsocalledinterferencerangeinmanyliteratures[ 91 ]sinceothertransmissionsinthisrangemayintroduceenoughinterferencetocorruptthein-tendedsignal.Thewidelyusednetworksimulationtoolns2implementsthesettingsofWaveLANcardfromLucentcompany.AndthedefaultvaluesareCPThresh=10dB,dt=250m;c1:78,ands2:2.Somerecentliteratures[ 104 102 ]aboutpowercontrolschemesadoptCPThresh=6dB,c1:41,ands2:2.Thus,itisreason-abletoassumethattheradiusofsensing/interferencerangecouldbemorethantwiceoftransmissionrange.

PAGE 181

Figure7: Asimplescenariotoillustratetheproblems tosomenode,resultinginacollisionatthereceivingnode.Fig. 7 illustratesasimpleexample,wherethesmallcirclesindicatetheedgesoftransmissionrangeandthelargecirclesindicatetheedgesofthesensingrange.DisthehiddenterminalofA.ItcannotsenseA'stransmissionbutmaystillinterferewithB'sreceptionifDbeginsatransmission. Anexposedterminalistheoneoutsideofthesensingrangeofthereceiverbutwithinthatofthetransmitter.Theexposednodesensesthemediumbusyanddoesnottransmitwhenthetransmittertransmits,leadingtobandwidthunder-utilization.InFig. 7 ,FistheexposedterminalofA.WhenAistransmittingtoB,FsensesA'stransmissionandkeepssilent.However,FcantransmittoothernodesoutsideofA'ssensingrangewithoutinterferingwithB'sreception. Inthefour-wayhandshakeprocedures,RTS/CTSandDATA/ACKarebidirectionalpacketsexchanged.Thereforetheexposednodeofoneofthetransmitter-receiverpairisalsothehiddennodeoftheother.Besidesthehiddenterminal,theexposedterminalofthetransmittershouldnotinitiateanynewtransmissioneitherduringtheongoingtransmissiontoavoidcollisionwiththeshortpacketsCTSorACK.Thisleadstosignicantinefciencyofthespatialreuse.

PAGE 182

carriedbyRTS/CTS/DATA/ACKnotiestheneighborstokeepsilentduringacertainperiodindicatedbytheNAVvalue. NAVsetupprocedurecannotworkproperlywhentherearecollisions.AsshowninFig 7 ,AwantstosendpacketstoB.TheyexchangeRTSandCTS.IfEistransmittingwhenBtransmitsCTStoA,BandE'stransmissionwillcollideatC,andCcannotsetitsNAVaccordingtothecorruptedCTSfromB. NAVsetupprocedureisredundantifanodeiscontinuouslysensingthecarrier.Forexample,inFig. 7 ,transmissionrangesofbothAandBarecoveredbythecommonareaoftheirsensingranges.Withoutcollisions,CcansetNAVcorrectlywhenreceivingB'sCTS.However,itcanalsosenseA'stransmissionwhichpreventsCfromtransmittingevenwhenthereisnoNAVsetupprocedure.RTS'sNAVisnotnecessaryeitherbecauseanynodewhichcanreceiveRTScorrectlycanalsosenseB'sCTSandsucceedingDATAandACK,andwillnotinitiatenewtransmissiontointerrupttheongoingtransmission. NAVsetupproceduredoesnotsolvethehiddenterminalproblemsevenifthereceivercancorrectlyreceiveCTSandsetitsNAV.InFig. 7 ,DisthehiddenterminalofAandoutoftransmissionrangeofB.ItcannotsenseA'stransmissionandcannotcorrectlyreceiveB'sCTSeither.Thus,whenAistransmittingalongdatapackettoB,Dmayinitiateanewtransmission,whichwillresultinacollisionatB.

PAGE 183

prioritythanthenewonebecauseitresetsitsbackoffwindowsizeandhasmuchshortervaluethanthatofnewone.Sotheoldsenderhasahighprobabilitytocontinuetotransmitandthenewonecontinuesdoublingthebackoffwindowsizeanddiscardspacketswhenthemaximumnumberoftransmissionattemptsisreached.Thiswillthereforeresultinseriousunfairnessamongowsandseverepacketdiscarding. Forexample,inFig. 7 ,whenDistransmittingtoE,AsendsRTStoBbutwillnotreceivetheintendedCTSfromB.ThisisbecauseBcannotcorrectlyreceiveA'sRTSduetocollisionfromD'stransmission.Then,Akeepsdoublingcontentionwindowandretransmittinguntilitdiscardsthepacket.IfDhasaburstoftrafctoE,itwillcontinuouslyoccupythechannelwhichwillstarvetheowfromAtoB. Thehiddenterminalproblemonlymakesthereceiverblockingproblemworse.Intheaboveexample,evenifAhasachancetotransmitapackettoB,itshiddenterminalDcouldstarttransmissionandcollidewithA'stransmissionatBbecauseDcannotsenseA'stransmission.Therefore,AalmosthasnochancetosuccessfullytransmitapackettoBwhenDhaspacketsdestinedtoE. Anotherabnormalityisthatpacketscontinuouslyaccumulateattherstfewhopsofthepath.Thereasonisthatthetransmissionattherstfewhopsencounterslesscontentionthanthatatsubsequentnodes.Onesimpleexample,asshowninFig. 7 ,isthechaintopologywithmorethan5hopswherenodesareseparatedbyxedlengthalittlelessthanthemaximumtransmissiondistance.Therstnodeisinterferedbythreesubsequentnodes.Thisnumberisfourforthesecondnodeand5forthethirdnode.Thismeanstherstnode

PAGE 184

Figure7: Chaintopology couldinjectmorepacketsintothechainthanthesubsequentnodescouldforward.Lietal.havediscussedthisphenomenainthepaper[ 91 ]andindicatedthat802.11MACfailstoachievetheoptimumthroughputforthechaintopology.

PAGE 185

Tosummarize,manyaforementionedproblemscannotbesolvedifasinglechannelisusedintheIEEE802.11MACprotocol. 122 59 ]isusedtosolvethehiddenterminalproblem.ACKisunnecessaryherebecauseourprotocolcanguaranteethatthereisnocollisiontoDATApackets.Todealwithwirelesschannelerrors,weintroduceaNACKsignalwhichisacontinuingbusytonesignalwhenthereceiverdeterminesthatthereceivedDATApacketiscorrupted.Thesenderwill

PAGE 186

Figure7: Proposedprotocol notmisinterpretthisNACKsignalsincetherearenootherreceiversinitssensingrangeandhencenointerferingNACKsignals.ItwillconcludethatthetransmissionissuccessfulifnoNACKsignalissensed. OurprotocolDUCHAadoptsthesametransmissionpowerandcapturethresholdCPThreshinbothcontrolandDATAchannels.AndthetransmissionpowerlevelforcorrectreceivingRXThreshisalsothesameforthetwochannelssothatthetwochannelshavethesametransmissionandsensingrange.ThebasicmessageexchangesequenceisshowninFig. 7 7 ,whenAnishestransmittingitsRTStoB,FshouldwaitatleastlongenoughforAtonishreceivingthepossibleCTS/NCTSfromB.CTS/NCTS

PAGE 187

DATAchannelisbusy,itreturnsNCTS.TheNCTSprovidestheestimatefortheremain-ingDATAtransmissiontimeinitsdurationeldaccordingtothedifferencebetweenthetransmissiontimeofmaximumDATApacketandthelengthithassensedabusymediumintheDATAchannel.DATA ThesenderassumesthatitsDATAtransmissionissuccessfulifthereisnoNACKsignalsensedduringtheNACKperiod.Otherwise,itassumesthatitstransmissionfailsbecauseofwirelesschannelerrorandthenstartstheretransmissionprocedure. Inaddition,duringtheNACKperiodbesidestheDATAtransmissionperiodanyothernodesinthesensingrangeofthesenderarenotallowedtobecomethereceiverofDATA

PAGE 188

packets,andanyothernodesinthesensingrangeofthereceiverarenotallowedtobecomethesenderofDATApackets.ThisistoavoidconfusionbetweenNACKsignalsandthenormalbusytonesignals. Intheabovemessageexchange,ourprotocoltransmitsorreceivespacketsinonlyonechannelatanytime.Weonlyusereceivebusytonesignalandnottransmitbusytonesignal.SoitisnecessarytosensetheDATAchannelbeforetransmittingCTS/NCTSpacketstoavoidbecomingareceiverinthesensingrangeofthetransmittersofsomeongoingDATApackettransmissions. 7 ,BbroadcastsbusytonesignalwhenitreceivesDATApacketfromA.ThehiddenterminalofA,i.e.,D,couldhearB'sbusytonesignalandthuswillnottransmitintheDATAchanneltoavoidinterferencewithB'sreception.Thus,thebusytonesignalfromtheDATA'sreceiverpreventsanyhiddenterminalsoftheintendedsenderfrominterferingwiththereception.Therefore,noDATApacketsaredroppedduetothehiddenterminalproblem.Solutiontotheexposedterminalproblem 7 ,BistheexposedterminalofDwhenDistransmittingDATApackettoE.BcouldinitiateRTS/CTSexchangewithAthoughitcansenseD'stransmissionintheDATAchannel.AftertheRTS/CTSexchangeissuccessfulbetweenBandA,BbeginstotransmitDATApackettoA.SinceAisoutofthesensingrangeofDandEisoutofsensingrangeofB,bothAandEcouldcorrectlyreceivetheDATApacketdestinedtothem.Thus,theexposedterminalcouldtransmitDATApacketsinDUCHAwhichcouldgreatlyenhancethespatialreuseratio.

PAGE 189

7 ,BistheblockedreceiverintheIEEE802.11MACprotocolwhenDistransmittingDATApacketstoE.InourprotocolDUCHA,BcancorrectlyreceiveA'sRTSinthecontrolchannelwhileDsendsDATApacketsintheDATAchannel.ThenBreturnsNCTStoAbecauseitsensesbusymediumintheDATAchannel.ThedurationeldofNCTScontainstheestimatefortheremainingbusyperiodintheDATAchannelwhichtakestonishD'stransmission.WhenAreceivestheNCTS,itdefersitstransmissionandstoptheunnecessaryretransmissions.ItretriesthetransmissionaftertheperiodindicatedinthedurationeldofNCTS.OncetheRTS/CTSexchangeissuccessfulbetweenAandB,AbeginstotransmitDATApackettoB.BwillcorrectlyreceivetheDATApacketbecausethereisnohiddenterminalproblemforreceivingDATApackets.Improvementofspatialreuse 7 ,DisthehiddenterminalofAwhenAistransmittingDATApackettoB.AftertheRTS/CTSexchangebetweenEandDissuccessfulinthecontrolchannel,EcouldtransmitDATApacketstoD.SinceDisoutofA'ssensingrangeandBisoutofE'ssensingrange,bothDandEcouldcorrectlyreceivetheintendedDATApackets.ThusDUCHAcouldgreatlyincreasespatialreusebyallowingmultipletransmittersormultiplereceiversinthesensingrangeofeachothertocommunicate.Atthesametime,therearenocollisionsforDATApacketsaswellastheNACKsignalsbecausethereisonlyonetransmitterinitsintendedreceiver'ssensingrangeandonlyonereceiverinitsintendedtransmitter'ssensingrange.Inherentmechanismtosolvetheintra-owcontentionproblem

PAGE 190

upstreamanddownstreamnodesinitssensingrangearepreventedfromtransmittingDATApacketsduringtheNACKperiod.Infact,thiscouldachieveoptimumpacketschedulingforchaintopologyanditissimilarforanysingleowscenario. Forexample,inFig. 7 ,node1hasthehighestprioritytoaccessthechannelwhenitreceivesonepacketfromnode0andhenceimmediatelyforwardsthepackettonode2.Forthesamereason,node2immediatelyforwardsthereceivedpackettonode3.Thennode3forwardsthereceivedpackettonode4.Becausenode0cansensenode1and2'stransmissions,itwillnotinterferewiththesetwonodes.Node0couldnotsendpacketstonode1eitherwhennode3forwardspacketto4becausenode1isintheinterferencerangeofnode3.Whennode4forwardspacketto5,node0couldhavechancetosendpackettonode1.Ingeneral,nodeswhichare4hopsawayfromeachotheralongthepathcouldsimultaneouslysendpacketstotheirnexthops.Thustheprocedurecouldutilize1/4ofthechannelbandwidth,themaximumthroughputwhichcanbeapproachedbythechaintopology[ 91 ]. ThereisnocollisionforNACKsignal,i.e.,thecontinuingbusytone,either,becausethereisonlyoneDATAreceiverinthesensingrangeofanyongoingsenderintheDATAchannel.AftersuccessfulRTS/CTSexchangebetweenthesenderanditsintendedreceiver,allothernodesinthesensingrangeofthesendercansenseitstransmissionintheDATAchannelandthusarerestrictedfrombecomingDATAreceivers. ThecontroloverheadcouldbereducedalthoughweintroduceanewNCTSpacketandanewNACKsignal.First,NCTSisonlytransmittedwhentheintendedreceivercannotreceiveDATApacket.ItcansavealotofunnecessaryretransmittedRTSpackets

PAGE 191

asdiscussedinSection 7.4.3 .Second,NACKsignaloccursonlywhentheDATApacketiscorruptedduetochannelfading,andhenceitstransmissionfrequencyisalsomuchsmallerthanthatofACKpacketsinthe802.11MACprotocol.Third,thereisnocollisionforDATApacketsandhencethetransmissionsofRTSandCTSforcorruptedDATApacketsaresaved. 7.5.1SimulationEnvironments 7 Table7: Defaultvaluesinthesimulations Preambleofallkindsofpackets 192s Controlchannelspeed 0.3Mbps Datachannelspeed 1.7Mbps DATAratein802.11 2.0Mbps Capturethreshold 10dB LengthofRTS 160bits LengthofCTS 112bits LengthofNCTS 112bits LengthofACK 112bits LengthofNACKsignal 150s DATAPacketsize 1000Bytes Inoursimulationstudy,severalimportantperformancemetricsareevaluated,whicharedescribedbelow:

PAGE 192

Figure7: Onesimpletopology efcientlyutilizedbytheapplicationsorthetrafc.Ifallowsareone-hopows,thisisthesameastheaggregatedend-to-endthroughput,referredtoastheaggre-gatedthroughputinthegures. ThecollidedDATApacketsandthediscardedDATApacketshavealsobeenevaluatedinsomecases.ThecollidedDATApacketsarethosetransmittedbutcorruptedbythehiddenterminals.ThediscardedDATApacketsarethosediscardedduetocontinuousfailedretransmissionsofRTSorDATApackets. 7 ,wheretherearehiddenterminals,exposedterminalsandreceiverblock-ingproblemsifIEEE802.11MACprotocolisused.Hiddenterminals

PAGE 193

(b) (c) (d) Simulationresultsforthesimpletopology

PAGE 194

Fig. 7(a) showsthatthenumberofcollidedDATApacketsincreaseswiththeofferedloadinIEEE802.11whileourprotocolhasnocollisionswiththeDATApackets.ThisinfactveriesthatthereisnohiddenterminalproblemforthetransmissionofDATApacketsinourprotocol.ThereasonisthatB'sbusytonesignalpreventsthehiddenterminalCfromtransmittingandhencethereisnocollisionatBandhenceBcanstillreceiveA'sDATApackets.However,intheIEEE802.11MACprotocol,ChasnowaytoknowthatAistransmittingDATApacketstoBandhencecausecollisionsatBifCbeginstransmissions.Exposedterminals InIEEE802.11MACprotocol,BandCcannottransmitDATApacketsatthesametimewhiletheycaninourDUCHAprotocol.Soourprotocolshouldhavemuchhigheraggregatedthroughputinthissimplescenariounderheavyofferedload.Theimprovementisabout55%asshowninFig. 7(b) .Receiverblockingproblem Fig 7(c) showsthatinIEEE802.11thesenderA,whoseintendedreceiverBisblocked,cannotsuccessfullytransmitanypackets.ThisisbecausethatBcouldnotcor-rectlyreceiveA'sRTSandthusAcontinuouslydiscardsDATApacketsaftermultipletrans-missionfailuresofRTSpackets.WhileinourprotocolDUCHA,thecontrolpacketsaretransmittedinaseparatechannelandtheblockedreceivercouldreturnanNCTSpackettoitsintendedsenderduringtheperiodofneighboringDATAtransmissions.Furthermore,inourprotocol,AcanobtainapartofthebandwidthtotransmitDATApacketswhilein

PAGE 195

IEEE802.11,A'sDATAtransmissionswillbecorruptedbyitshiddenterminalCeveniftheRTS-CTSexchangeissuccessfulbetweenAandB.Improvementofspatialreuse Fig 7(d) showsthatourprotocolhasmuchhigheraggregatedthroughputthanIEEE802.11MAC.ThelattersuffersnotonlyfromthepoorspatialreusebutalsofromthecollisionsamongRTS,CTS,DATAandACKpacketssinceBandCarehiddenterminalsofAandD,respectively.Intra-owcontention 7.4 .Fig. 7 showstheaggregatedthroughputofa9-nodechaintopology.DUCHAimprovesthemaximumthroughputbyabout25%andhasa40%higherthroughputthanIEEE802.11MACunderheavyofferedload.ThisisbecauseDUCHAhasalargespatialreuseratiointheDATAchannelandcouldachievetheoptimumpacketschedulingforthechaintopologyindependentofthetrafcloadwhileIEEE802.11MACsuffersfromcollisionsunderheavyload. WeobservefromFig. 7 thattheaggregatedthroughputforallowsdecreaseswhentheminimumsource-destinationdistanceincreases.TheaggregatedthroughputofourprotocolishigherthanthatofIEEE802.11MAC.Anditdegradesmuchslowerin

PAGE 196

Figure7: End-to-endthroughputforthe9-nodechaintopology Figure7: Simulationresultsforrandomone-hopowswithdifferentminimumonehopdistance

PAGE 197

ourprotocolthaninIEEE802.11MACanditisimprovedbyabout8%to28%whentheminimumsource-destinationdistanceincreasesfrom0mto200m. Thisisreasonable.Forexample,AandBarethesource-destinationpair.ThelargerthedistancebetweenAandB,thelargerthehiddenareawherenodescannotsenseA'stransmissionbutcansenseB'stransmission.SoinIEEE802.11MAC,thehiddenterminalproblembecomesmoreseverewhenthedistancebetweenAandBbecomeslarger.Ontheotherhand,inIEEE802.11MAC,allthenodesinthesensingrangeofAorBshouldnottransmitanything,i.e.,bothsensingrangesofAandBcouldnotbereusedbyothertransmissions.However,inourprotocolDUCHA,theexposedarea,wherenodescansensethesender'stransmissionbutnotthereceiver'stransmission,couldbereusedfornewsenders,andthehiddenarea,wherenodescansensethereceiver'stransmissionbutnotthesender'stransmission,couldbereusedfornewreceivers.Thusthelargerthesource-destinationdistanceis,thehigherthesystemcapacityourprotocolDUCHAcouldobtainthantheIEEE802.11MAC. Infact,mostofthecurrentroutingalgorithmsmaximizethedistancebetweentheupstreamnodeandthedownstreamnodewhenselectingapathtoreducethehop-count,thedelayandthepowerconsumptionfordeliveringthepacketsfromthesourcetothedestination.OurprotocolDUCHAalsogivesagoodsolutiontotheintra-owcontentionproblemandcouldachieveoptimumpacketschedulingforthechaintopology.

PAGE 198

(b) (c) (d) Simulationresultsformultihopowsinrandomtopology

PAGE 199

7(a) thatwhentheminimumhop-countforeachowin-creases,theaggregatedend-to-endthroughputofbothprotocolsdecreases.Thisisreason-ablebecausepacketsofmultihopowshavetopassmorelinksandthusconsumemoreresourceforthesamearrivingtrafc. ThethroughputofIEEE802.11MACreducesmoredramaticallythanthatofDUCHAwhentheminimumhop-countforeachowincreases.TheimprovementofthroughputcomparingtotheIEEE802.11MACisabout2and5times,respectively,forthescenarioswheretheminimumofthehop-countsforallowsare3and5.AggregatedOne-HopThroughput 7(b) .ThisimpliesthatDUCHAcouldeffectivelyutilizemuchmoreresourceofthewirelessadhocnetworksthanIEEE802.11MACdoes. TheresourceefcientlyutilizedbytheowsgreatlydecreasesinIEEE802.11MACwhenthehopcountofeachowincreases,whileourprotocolDUCHAmaintainsarela-tivelyhighresourceutilizationratioformultihopowswithdifferenthopcounts.Andourprotocolevenefcientlyutilizemoreresourcewhenthehopcountforeachowincreases.ThisimpliesthatIEEE802.11MACisnotappropriateformultihopadhocnetworkswhileourprotocolDUCHAworkswellandisscalableforlargernetworkswheretheowshavelargerhopcounts.TransmissionEfciencyofDATAPackets 7(c) Inaddition,ourprotocolmaintainsrelativelystabletransmissionefciencyofDATApacketsforowswithdifferenthopcountswhiletheIEEE802.11MACdegradessigni-cantlywhenthehopcountforeachowincreases.ThereasonisthatourprotocolDUCHA

PAGE 200

notonlyhasnocollidedDATApackets,butalsohasmuchlessaccumulatedanddiscardedpacketsattheintermediatenodesalongthepaths.ThismeansthatourprotocolcouldsavesignicantresourceandlowerthepowerconsumptiontodeliverthesameamountofDATApackets.NormalizedControlOverhead 7(d) ,weobservethatthenormalizedcontroloverheadisalsomuchlessinourprotocolthanthatintheIEEE802.11MAC.ItlinearlyincreaseswiththeofferedloadforthemultihopowsintheIEEE802.11MACwhileourprotocolDUCHAmaintainsasmallstablevalue.Moreover,similartotheotherperformancemetrics,thenormalizedcontroloverheadmaintainsarelativelystablevalueforowswithdifferenthopcountsinourprotocolDUCHAwhileinIEEE802.11MACitbecomeslargerandlargerwhenthehopcountforeachowincreases.ThisimpliesthatourprotocolhasmuchhigherefciencytotransmitDATApackets.AndIEEE802.11MACdoesnotworkwellformultihopowsespeciallyunderheavyloadandwillresultintheexplosionofcontrolpackets,leadingtomorecontrolpacketsandlowerthroughput.

PAGE 201

theroutinglayerwithmuchlessunnecessaryreroutingrequestsbyprovidingmoreaccuratenext-hopinformation. Extensivesimulationsshowthatourprotocolimprovesthethroughputby8%-28%foronehopowsandbyseveraltimesforthemultihopowswhenitusesthesameto-talbandwidthasthatoftheIEEE802.11MAC.Inaddition,ourprotocolisscalableforlargenetworksandmaintainshighresourceutilizationratioandstablenormalizedcontroloverheadwhiletheIEEE802.11MACdoesnotworkwellformultihopowsunderheavytrafc.

PAGE 202

Inwirelessmultihopadhocnetworks,collisionduetohiddenterminalproblemiscommonandmakesitdifculttoprovidetherequiredqualityofserviceformultimediaservicesorsupportpriority-basedservices.Inthischapter,werstanalyzetheshortcom-ingsofexistingapproachestoalleviatethehiddenterminalproblem.Thenweproposeanewschemeinwhichthereceiversendsshortbusyadvertisementsoverthesamechanneltocleartheoorforreceiving.Carriersensingrangeissetassmallasaninterferencerangetoalleviatetheexposedterminalproblem.Thenewschemeonlyrequiresasingletransceiverandasinglechannel.Weanalyzeandevaluatetheperformanceoftheproposedschemeextensively.Theresultsshowthatthenewschemehasamuchhigherefciencythantheexistingapproachesusingasinglechannelandasingletransceiver. Inthehiddenterminalproblem,packetcollisionhappensattheintendedreceiverifthereisatransmissionfromahiddenterminal.Here,ahiddenterminalisanodethatcannotsensetheongoingtransmissionbutisabletointroduceenoughinterferencetocorruptthereceivingifittransmits.ForexampleinFig. 8 ,thereisanongoingtransmissionfromA 183

PAGE 203

Figure8: Hiddenterminalproblem toB.CisahiddenterminalofAandmaytransmitduringtheongoingtransmissionfromA,whichleadstocollisionatB.BecauseCdoesnotknowwhetherAistransmittingornot,itcanoccupythechannelatanytimeandthequalityoftheowfromAtoBcannotbeguaranteedwheneverthereareanypacketsfromCtoD.Wewillillustratemoredetailsoftherelatedcarriersensing,transmissionandinterferencerangesinSection 8.2 Awidelystudiedsolutiontothehiddenterminalproblemistheout-of-bandbusytoneapproach([ 122 59 133 165 163 161 153 ]andreferencestherein).ReceiversendsoutthebusytonesignalonthebusytonechannelwhilereceivingDATApacketsontheDATAchannel.Allnodesinthenetworkarerequiredtomonitorthebusytonechannel.Ifanodeoverhearsthebusytonesignal,itmustkeepsilenttoavoidpossiblecollision.Thisapproachcanwelladdressthehiddenterminalproblem,butitrequiresbothanadditionalchannelandanadditionaltransceiver. Severalapproaches([ 138 142 161 153 155 ]andreferencestherein)havealsobeenproposedtoaddressthehiddenterminalproblemwithouttherequirementoftheadditionalchannelandtransceiver.Acommonapproachistousealargecarriersensingrange(LCS)tocovertheinterferencerangearoundthereceiver,asshownintheleftpartofFig. 8 .Ifthereisnoobstructinbetween,allthenodeswhosetransmissionscaninterferewiththepacketreceptioncansensethetransmissionfromthetransmitterandhencearerequiredtokeepsilenttoavoidcollision.However,thisapproachdecreasesthespatialreuseratio

PAGE 204

Figure8: CarriersensingrangeandinterferencerangeinLCSandSBA-MAC bysilencingalotofnodesthatareoutoftheinterferencerangeofthereceiveranddonotinterferewiththeongoingtransmissionandreceptioniftheytransmit.Furthermore,itdoesnotcompletelysolvethehiddenterminalproblemifthereisobstructbetweenthenodes.Forexample,inFig. 8 ,nodeCcannotsensethetransmissionfromAforanobstructinbetweenandisstillahiddenterminal.Avariationoftheapproachistomaintainthesamecarriersensingrangebuttoreducethetransmissionrangebyenforcingahigherpowerthresholdforpacketreception[ 138 ].Thebasicideaisstilltocovertheinterferencerangeofthereceiverwithinthecarriersensingrangeofthetransmitter.Itsharesthesamespatialreuseratioanddoesnotaddressthehiddenterminalproblemeitherwhenobstructionexists. Toaddresstheshortcomingsoftheaboveapproaches,ahiddenterminalhastodefertheirtransmissionsaccordingtoareceivedorsensedsignal/packetfromthecurrentreceiveronthesamechannelforDATAtransmission.FullmerandGarcia-Luna-AcevesproposedaschemeFAMAinthepaper[ 47 ]touseaCTSdominancemechanismtoensurecollision-freedatapacketreception.Thismechanismrequiresnodessensinganynoisetodefertheirtransmissionlongenoughforamaximum-lengthdatapackettobereceived.ItwillmistakenlytreatcollisionsandanyundecodabletransmissionsofframesotherthanCTStobeCTSdominanceeandthenwastethechannelinthelongdeferringtime.YehproposedCSMA/IBinthepaper[ 144 ]torequirethereceivertotransmitashortsignalorinbandbusytonebetweenthereceiveddatafragments.Anynodesoverhearingthesignalhaveto

PAGE 205

defertheirtransmissionsforadurationequaltothetransmissiontimeforamaximumdatafragment.ComparedtoFAMA,CSMA/IBcanreducethedeferringtimesignicantlyifthelengthofamaximumdatafragmentismuchlessthanamaximum-lengthdatapacket.However,busytoneperiodsincreasethetotaltransmissiontimeofadatapacket.Datafragmentsalsointroducemorecontroloverhead,likethephysicalandMAClayerheaders[ 68 ].TheperformanceofCSMA/IBhasnotbeenwellevaluated.Howtosetthelengthofbusytonesignalandmaximumdatafragment,andwhatistheirimpactontheperformancedeservecarefulstudies. Inthischapter[ 159 ],weproposeanewMACschemeusingdummybitsandshortbusyadvertisement(SBA)signalsbasedontheCSMA/CA(carriersensemultipleac-cesswithcollisionavoidance)ortheIEEE802.11MACscheme.InthebasicSBA-MACscheme,severalshortperiodsofdummybitsareinsertedintheDATAframe.Duringtheseperiods,thereceiverswitchestothetransmissionmodeandtransmitashortbusyadver-tisementconsistingofsynchronizationsymbols,andthenswitchesbacktocontinuethepacketreception.AnodedefersitstransmissionforaBIFS(interframespacingduetobusyadvertisement)periodafterdetectingaSBAsignaloranynoise.TheaboveSBApro-cedureisonlyusedwhenahiddenterminalisdetectedandthenormal802.11operationisusedotherwise.Inaddition,thetransmissionpowerofbusyadvertisementiscontrolledtoimprovespatialreuseratiowhilestillsolvingthehiddenterminalproblem.WeanalyzetheperformanceoftheSBA-MACschemeandstudytheimpactofthelengthofBIFSonthesystemperformance.TheresultsshowthatSBA-MACcansignicantlyoutperformthepreviousapproachesusingasinglechannel.ThefeaturesandadvantagesofSBA-MACaresummarizedasfollows.

PAGE 206

Therestofthischapterisorganizedasfollows.Section 8.2 illustratesvariousrangesincarriersenseMACprotocols.TheSBA-MACschemeisproposedinSection 8.3 .InSection 8.4 ,westudyhowtocontroltransmissionpowertofurtherincreasespatialreuseratio.ThenweanalyzeandevaluatetheperformanceoftheproposedschemeinSection 8.5 .Finally,weconcludethechapterinSection 8.6

PAGE 208

receivedinterferencepowerfromonenodeislargerthanPi,thatnodeisintheinterferencerangeoftheconsiderednode;and wherePristhereceivedpoweroftheintendedpacketatthecurrentreceiver.Ifweassumethateverynodeusesthesametransmissionpower,wehave wherediistheradiusoftheinterferencerange,dhisthedistancebetweenthecurrenttransmitteranditsintendedreceiverorsimplyhopdistanceofthecurrenthop/link,diistheradiusofthetypicalinterferencerange,isthepathlossexponent,anddtisthemaximumcommunicationdistance. Noticethattwoormoreconcurrentinterferencesignalsmayexist.TheinterferencerangedenedbyPiinEquation( 8.1 )isnotlargeenoughtocovernodeswhichmaycorruptthepacketreception.Normally([ 155 ]),PiandPishouldbe2or3dBsmallertoaddressthisissue. SNRisusuallyrequiredtobelargerthan0dBforcorrectreception.Therefore,Pi>Pseandtypicalinterferencerangeislargerthancommunicationrange.Inthedefaultsettingsinthewidelyusedsimulationtoolns2,carriersensingradiusis2.2timesoftrans-missionradius,andinterferenceradiusisabout1.78timesoftransmissionradiuswhencapturethresholdissetas10dB.

PAGE 209

Figure8: Four-wayhandshakewithbusyadvertisementsignals constructashortbusyadvertisementsignalandhowtosetvariousparametersintheSBA-MACscheme.Finally,wediscusswhythisschemecangreatlyincreasethespatialreuseratioandthecompatibilityissuewiththelegacy802.11nodes. 8.3.7 WhentheSBAprocedureisadopted,atransmitterdividesthepayloadoftheDATAframeintoseveralpartsorfragmentsandinsertsasmallblockofbitsbetweentwoadjacentparts.Thesebitsaredummybitsandcanbeequaltoanyvalues.Atransmissionperiodofthesedummybitsisreferredtoasanintra-data-framespacingoraninter-data-fragmentspacing(IDFS). DuringanIDFSperiod,theintendedreceiverignoresthereceivedsignal,sendsoutabusyadvertisementsignaloverthesamechanneltonotifythehiddenterminalsoftheongoingtransmission,andthenswitchesbacktocontinuethepacketreception.Thecorre-spondingmessagesequenceisshowninFig. 8 Toprotectthedatafragments,anydevicesensingthebusyadvertisementsignalinthetypicalinterferencerange(i.e.thesensedpowerislargerthanPi,refertoSection 8.2 )shouldkeepsilentforacertainperiod.WerefertothisperiodasaBIFSperiod,oran

PAGE 210

8.2 )asshownintherightpartofFig. 8 ,i.e. AnynodesensesasignalwithapowerlevellargerthanthetypicalinterferencepowerPishouldindicateabusychannelanddeferitstransmissionduringthebusyperiod.Ifthesignalisundecodableorabusyadvertisementsignal,itisfurtherrequiredtokeepsilentforatleastaBIFSperiodafterthesignalisnished.IfthesignalisacorrectlyreceivedMACframe,itkeepssilentduringtheperiodindicatedinthedurationeldintheframeusingtheoriginalvirtualcarriersensingmechanism. Inthisway,atransmitteronlysilencethenodesthatitmayinterferewithandthatmayinterferewithitsreceptionofanACKframe.Abusyadvertisementsignalonlysilencesthosehiddenterminalswhichcaninterferewiththecurrentpacketreception.Inthischap-terwealsorefertotheinterferencerangearoundareceiver,whichissilencedbyabusyadvertisement,asabusyadvertisementrangehereafter.Therefore,ourapproachallowsmoreconcurrenttransmissionsandhencecansignicantlyincreasethespatialreuseratiocomparedtotheapproachusingalargecarriersensingrangeasshowninFig. 8

PAGE 211

andtoswitchbackforreceiving. 68 71 69 70 ]. LetTBIFSbethelengthofaBIFSperiod.NoticethatanEIFSprocedureisalreadyadoptedintheIEEE802.11MACprotocol.AnodeisrequiredtokeepsilentforatleastanEIFSperiodafteritdetectsanundecodablesignal.TheEIFSperiodisusedtoprotectthereceptionofanACKframe.Toprovidethesamefunction,theBIFSprocedurereplacestheoriginalEIFSprocedureandTBIFSshouldbelargerthanorequaltoTEIFS.SinceanodeonlyknowsthatitistheintendedreceiverafterreceivingthephysicalandMACheaders,BIFSshouldbelargeenoughtoprotectthereceptionoftheseheaders.So whereTPHYandTMACarethetransmissiontimeforthephysicalheaderandtheMACheader,respectively.Tpropisthemaximumpropagationdelaybetweentwocommunicating

PAGE 212

Figure8: PositionsofIDFSperiodsintheDATAframe nodes.Ontheotherhand,TBIFSmustbelargerthanorequaltothemaximumtransmissiontimeofadatafragment.Noticethatbetweentwoconsecutivebusyadvertisementsignals,thereceiverneedstoconsumetimeTTRandTRTbesidesthetimeforreceptionofaDATAfragment.Therefore,tokeepahiddenterminalsilent,thetransmissiontimeforaDATAfragmentTfragmustsatify ifitisnotthelastfragment,and otherwise.Accordingly,atransmitterdividesadataframeintooneormorefragmentandplacessomedummybitsinbetweenthroughthefollowingmethod. IfRTS/CTSareused,atransmitterusesthefollowingproceduretoplaceIDFSperiodsintheDATAframe.

PAGE 213

8 showssuchaDATAframewithfourfragments. IfRTS/CTSarenotused,atransmitterusesthefollowingproceduretoplaceIDFSperiodsintheDATAframe. 8 showssuchaDATAframewithfourfragments. Aboveprocedureattemptstotransmitthelastbusyadvertisementatatimeasearlieraspossiblethantheendofthetransmission,andstilllateenoughtoprotectthereceivingofthelastdatafragment.WithanappropriatevalueofTBIFS,itisrarethatadeviceis

PAGE 214

requiredtokeepsilentaftertheongoingtransmissionisnishedduetoonesensedbusyadvertisement.WewillstudyhowtosetthevalueofTBIFSinSection 8.5 8.2 ,interferencerangechangeswiththereceivedpoweroftheintendedsignal.Ontheotherhand,thetransmissionpowerofbusyadvertisementcontrolsthesizeofthereservedareaaroundthereceiverwherenodesdefertheirtransmissionifoverhearingabusyadvertisement.Therefore,wecanadjustthetransmissionpowerofbusyadvertisementtoobtainareservedareaequaltotheareaoftheinterferencerange,whichisnormallysmallerthanatypicalinterferencerange.Wewillderivetheappropriatetransmissionpowerofashortbusyadvertisementsignalasfollows. InSBA-MAC,carriersensingthresholdPcsisalsousedtodeterminetheedgeoftheinterferencerange.Thatistosay,whenthesensedpowerofabusyadvertisementsignalislessthanPcs,anodedeterminesthatitisoutsideoftheinterferencerange,otherwiseintherange.LetPtdenotethetransmissionpowerofbusyadvertisement,whichresultsinatypicalinterferencerange.SupposePtisalsousedtotransmitotherMACframesliketheDATAframe.PrdenotesthereceivedpowerofaDATAframeatthereceiver.Psedenotesthepowerdenedbythereceiversensitivity.P0tbadenotesthetransmissionpowerofabusyadvertisementsignalthatdenesaninterferencerangesubjecttoPr.P0idenotesthereceivedpowerattheconsideredreceiverduetoanothertransmissionattheedgeoftheinterferencerange.Then

PAGE 215

TherstequationinEquation( 8.8 )comesfromthefactthatthepathlossfromtheconsid-eredreceivertoanodeattheedgeofthereservedareaisequaltothatfromanodeattheedgeofthereservedareatotheconsideredreceiver.ThesecondequationinEquation( 8.8 )indicatesthatSNRshouldnotbesacricedduetoasmallreservedareacomparedtothetypicalinterferencerangeforPr=Pse.Then Letdhdenotethedistancebetweenthetransmitteranditsintendedreceiver,dtdenotethemaximumtransmissiondistancedenedbythereceiversensitivity,dbdenotetheradiusofthetypicalinterferencerange,d0bdenotetheradiusoftheinterferencerange. Therefore,bydecreasingthetransmissionpowerofbusyadvertisementfromPttoP0tba,thebusyadvertisementrangeisreducetotheinterferencerangeandmaybemuchsmallerthanthetypicalinterferencerange.Inthisway,moreconcurrenttransmissionsareallowedinthenetworkandspatialreuseratioisincreased. Whenthecarriersensingrangearoundthetransmittercoversthewholeareaoftheinterferencerangearoundthereceiver(orbusyadvertisementrange),i.e., wecanchoosenottosendbusyadvertisementtoreduceoverheadifweknowthereisnoobstructbetweenthetransmitterandnodesintheinterferencerangeofthereceiver. 68 ]isusedtoindicatewhetherornottostarttheSBAprocedure.WerefertothisbitasanSBA-bit.IfanSBA-bitissetasoneinaDATAframe,itmeansthatSBAprocedureisused,andnot

PAGE 216

otherwise.IfanSBA-bitissetinanACKframe,itmeansthatSBAprocedureshouldbeusedinthenextDATAframetothisreceiver,andnototherwise. Normally,SBAprocedureisdisabled.OncethereisnoacknowledgementforasentDATAframe,thetransmitterassumesthereisahiddenterminalforthecurrenttransmissionandadoptsSBAprocedureforsubsequent(re)transmissionstothecurrentreceiver.WhenareceiverrstreceivesaDATAframewithSBA-bitsetasone,itwillsettheSBA-bitasoneinalltherespondingACKframesfromthenon.Onlyafteracertaintime,duringwhichthereisnosensedundecodablesignal,itassumesthatthereisnohiddenterminalandshoulddisabletheSBA-bitinrespondingACKframe.WhenatransmitterreceivesanACKframewiththeSBA-bitnotsetasone,itstopstheSBAprocedureifusedforthecorrespondingreceiver. TodifferentiateerrorsinaDATAframeduetointerferencefromahiddenterminalandrandomchannelbiterror,anewtypeofMACframe,sayNACK(negativeACK),isused.AslongasthemeasuredSNRatthephysicallayerislargerthanthenominalSNRrequirement,anodeassumesthatthereisnohiddenterminalandreturnanNACKifthereisanerrorinthereceivedDATAframe.Otherwise,anodedoesnotrespondanythingtothetransmitterforareceivederroneousDATAframe.TheproceduretosettheSBA-bitinanNACKframeisthesameasthatforanACKframe.WhenthetransmitterreceivesanNACKframe,itassumesthereisanerrorinthetransmissionandtheerrorresultsfromarandomchannelbiterror.ItadoptstheSBAprocedureforsubsequenttransmissionsoffragmentsaccordingtotheSBA-bitintheNACKframe. Intheworstcasewherethenoiseoorisveryhighandhencechannelbiterrorhappensfrequently,SBAprocedurewillbeadoptedforallDATAframetransmissions.AlthoughitnoticeablyincreasestheoverheadorthetransmissiontimeofaDATAframe,theimprove-mentduetoincreasedspatialreuseratioasdiscussedinSection 8.3.2 isstillhighenoughtoaccountforit.WewillstudytheperformancebyconsideringthesetwofactorsinSection 8.5

PAGE 217

AtransmittercanchoosenottosendthedummybitsduringIDFSperiodstosavealittleenergyaslongasthesilentperiodsdonotresultinthelossofsynchronizationinformationforthesubsequentdatafragmentstobereceivedatthereceiver. Accumulativeacknowledgementisspeciallyusefulwhenthechannelbiterrorratioishigh.Thewholedataframemayhaveahighprobabilitytohaveanerror.However,eachdatafragmenthasamuchhigherprobabilitytobecorrectlyreceived.Therefore,accumulativeacknowledgementcansavealotofretransmissioncostinthiscase.

PAGE 218

68 ],theheadpartofthedataframemaycollidewiththetailpartoftheRTSframe.Toavoidthistypeofhiddenterminalproblemandensureanerror-freedatareception,anewCTSframelongerthanaRTSframeshouldbeusedlikethatintheFAMAscheme[ 47 ].Thusintheabovescenario,thehiddenterminalcansensethetailpartofaCTSframeandhencedeferitsowntransmissionaccordingtothecarriersensingmechanism.

PAGE 219

Figure8: PowercontrolinSBA-MAC Inatypicalscenario,dh=dt.ThetransmissionpowerofDATAframeisPt,thereceiverpoweratthereceiverisPse,carriersensingthresholdisPcs,carriersensingradiusisdcs,busyadvertisementthresholdisPba=Pcs=Pi,radiusoftypicalinterferencerangeisdb=dcs=di.TransmissionpowerofabusyadvertisementsignalisalsoPt. Whenthetransmitter-receiverdistancedh
PAGE 220

sensedpoweratdbandd0bawayarePcsandP0i,respectively.Theseleadto Becausepathlossbetweenapairoftwolocationsdoesnotchangewiththetransmissionpower,wehave AccordingtoEquation( 8.13 ) NoticethatwiththesametransmissionpowerPt,thereceivedpowerarePseandPratdistancedtanddh,respectively, andaccordingtoEquation( 8.12 ),( 8.13 )and( 8.14 ),wehave BecauseP0t6PtandP0tba6Pt,d0cs6dcsandd0b6db.So Apparently,whenx=1,thereisonlypowercontrolforbusyadvertisement,d0cs=dcsandd0b=dcsdh 8.10 );andwhenx=dh Toincreasethespatialreuse,weneedminimizetheareaS(d0cs;d0b)coveredbythecarriersensingrangearoundthetransmitterandthebusyadvertisementrangearoundthe

PAGE 221

receiver.ItiseasytoshowthatS(r1;r2)isdenedby NowtheproblembecomestominimizeS(d0cs;d0b)underconditionsEquation( 8.15 ),( 8.17 )and( 8.18 ).Thatistosay Itisnotdifculttoprovethatwhend0cs=d0b=dcsq IfPr>Pse,thetransmissionpowerofDATAframecanbereduced.Whentransmis-sionpowerisP0t,carriersensingrangewitharadiusofd0csaroundthetransmittershouldstilljustcovertheinterferencerangewitharadiusofd0iaroundthereceiver.Pcsisstillthe

PAGE 222

sensedpowerattheedgeofthereducedcarriersensingrange. Letbethepathlossexponent.ThetransmissionpowerisreducedfromPttoP0t,sotheradiusofthecarriersensingrangeisalsoreducedfromdcstod0cs.BecausethesensedpowerattheedgeofthecarriersensingrangedoesnotchangeandisstillequaltoPcs,wehave IfanothernodetransmitsattheedgeoftheinterferencerangearoundthereceiverwithpowerPt,thepoweroftheinterferencereceivedattheconsideredreceiverisP0iandsatis-es Withasmallercarriersensingrangeandasmallerinterferencerange,weshouldstillmain-tainatleastthesamesignaltointerferenceratioasthatinthetypicalscenario: Thoughthetransmissionpowerisreduced,thepathlossshouldbethesame: AccordingtoEquation( 8.24 ),( 8.25 ),( 8.26 )and( 8.27 ), Replacingd0iwithdcs0dh,wecansolved0csbynoticingthatd0cs>0:

PAGE 223

Whendh=dt,fromEquation( 8.29 )and( 8.22 ),d0cs=dcswhichisthedesiredresultforthetypicalcase.Nowwiththeknowledgeofd0cs,wecandeterminethenewtransmissionpowerP0tbyEquation( 8.24 ). 8.5.1SpatialReuseRatio Whenpowercontrolisusedfordataframes, IntheapproachusingalargecarriersensingrangeasshownintheleftsideofFig. 8 ,theareaoccupiedbyeachtransmissionisSlcsand Whenpowercontrolisusedfordataframes, InSBA-MAC,thetransmissiontimeofadatapacketisincreasedduetoinserteddummybits.LetNbabethenumberofinsertedIDFSperiods.NoticethatthechanneltimeTpbausedforeachdatapacketalsoincludesthebackoffperiodandthedeferringtimeTBIFSduetosensedundecodablesignals,whichisalsoincreasedifBIFSislongerthan

PAGE 224

EIFS.LetpbabetheprobabilitythatanodenearthecurrenttransmissiondefersTBIFS,successfullycontendsforthechannelandbeginstotransmitafterthecurrenttransmissionisnished.Then1pbaistheprobabilitythatthetransmissionopportunityisobtainedbythecurrenttransmitterorreceiver,oranyanothernodethatcorrectlyoverhearsthecurrenttransmission.LetphbetheprobabilitythatatransmitterdeterminesthereisahiddenterminalandadoptstheSBAprocedure.Now,wecanobtain AccordingtotheprocedureinSection 8.3.4 ,wehave Itiseasytoshowthatwhen IntheFAMAscheme[ 47 ],weassumethatFAMAusesthesamecarriersensingrangetoobtaingoodspatialreuseratioasinSBA-MAC.Similarlywithpba,wedeneitaspfamainFAMA.ThechanneltimeTpfamausedforeachdatapacketinFAMAis

PAGE 225

Similarly,thechanneltimeTplcsusedforeachdatapacketintheapproachusingalargecarriersensingrangeisequalto SinceSlcs>Sba,therearemorenodesthatsenseundecodablesignals.Thatistosay, NowwecancalculatethegainofSBA-MACcomparedwiththeapproachusingalargecarriersensingrangeis Fig. 8 showsSBA-MACcangreatlyreducetheoccupiedareabyeachtransmissioncomparedtoLCS.Thegainisfrom58%to80%withpowercontrolfordataframes,andfrom80%to144%withoutpowercontrolfordataframes.Powercontrolfordataframescangreatlyincreasethespatialreuse. Fig. 8 showstheoccupiedareabyeachtransmissionincreasesalongwiththein-terferenceradius,whichisdeterminedbytheSNRrequirement(Section 8.2 ).SBA-MACalwayshasasignicantgaincomparedtoLCS,whichisfrom31%to141%.

PAGE 226

Figure8: Occupiedareaforatransmissionnormalizedoverthecommunicationradius(PC:powercontrolforDATAframes) Figure8: Occupiedareaforatransmissionnormalizedoverthecommunicationradiuswhendh=dt

PAGE 227

Figure8: Channeltimeforatransmittedpacket Fig. 8 showsthechanneltimeforatransmitteddatapacketinSBA-MAC.WecanseethatTpbaonlychangesbyupto1.9%whenTBIFSisfrom364sto964salthoughthereisapparentlyanoptimalvalueofTBIFS. Fig. 8 illustratesthatSBA-MAConlyincreasesthechanneltimeforeachpacketbyabout0.85%to8.5%.WeplottheperformancegainKsbalcsandKsbafamaintheFigure 8 .ItdemonstratesthatSBA-MACcanimprovethethroughputby44%to53%whenTDATAisfrom10to1mscomparedtotheapproachusingalargecarriersensingrange.Theimprovementisabout68%to344%comparedtotheFAMAscheme.

PAGE 228

Figure8: Channeltimeforatransmittedpacket Figure8: PerformancegainofSBA-MACcomparedtotheapproachusingalargecarriersensingrangeandtheFAMAscheme

PAGE 229

terminalproblem.TheperformanceresultsshowthatSBA-MACnoticeablyoutperformstheexistingapproachesaddressingthehiddenterminalproblem.

PAGE 230

Alongwiththegrowingpopularityofsensorandadhocnetworks,variouskindsofservicesareexpectedtobesupported.Inwirelessadhocnetworks,thereareincreasingdemandsforwebtrafc,voiceoverIPandstreamingvideofromandtotheInternetviatheaccesspoints.Insensornetworks,event-drivenorperiodicallymonitoringservicesarecommon.However,variouslengthsofpacketsareusedindifferentservices.ShortpacketshaverelativelylargeoverheadattheMAC(mediumaccesscontrol)andphysicallayersandhencecansignicantlydecreasethenetworkthroughput.Inthischapter,weanalyzetheperformanceofadistributedadaptivepacketconcatenation(APC)schemewhichisproposedtoimprovethenetworkthroughput.TheAPCschemeworksattheinterfacequeueofthedatalinklayer.ItadaptivelyconcatenatesseveralshortpacketswhicharedestinedtothesamenexthopintoalongpacketforMAClayer'stransmissionaccordingtothecongestionstatusaswellastheobservedchannelstatus.Thetheoreticalanalysisisconductedinbothsinglehopnetworksandmultihopnetworks,andtheresultshowsthattheAPCschemecanincreasethethroughputbyupto4to16times. 211

PAGE 231

bandwidth,delayandpacketlength,whichprovidegreatchallengesfornetworkprotocolstoworkefciently. ShortdatapacketsoccupyarelativelargechannelresourceduetothexedphysicalandMAClayers'overhead.TheyalsoleadtocongestionsandsevereMACcontentionsmoreeasilythanlongdatapacketsgivenacertainamountofdatatrafc.FortheIEEE802.11protocols,thephysicallayeroverheadincludesapreamble,whichisusedtosyn-chronizethetransmitterandthereceiver,andsomecontroleldstonotifythereceiverofthechannelcodingandmodulationschemes.TheMAClayeroverheadincludesseveralMAClayercontrolframesconsistingofRTS(readytosend),CTS(cleartosend)andACK(acknowledge),MACaddressoftheDATAframesandinterframespacings,suchasSIFSandDIFS.TheshorterthepayloadoftheDATAframe,thesmallerthethroughputandthemorethewastedchannelresource. Severalschemes([ 80 65 113 76 ])havebeenproposedtoefcientlyutilizethetime-varyingchannelinwirelessLANswherenodescandirectlycommunicatewitheachother.Whenthechannelqualityisgood,severalpacketsaretransmittedbacktobackwithalargechannelrateatatime.Otherwise,asinglepacketistransmittedwithasmallchannelrate.Theseschemesareefcientinreducingtherelativeprotocoloverheadwhenalargechannelrateisused. Insensorandadhocnetworks,datapacketsoftenneedtobeforwardedseveraltimesbeforetheyreachthedestinations.Eachforwardingnodeneedstocontendforthechannelwithothernodesbeforeitcantransmitapacket.TheMAClayercontentionbecomesmoreseverewhencongestionhappensandalotofbackloggedpacketskeepnodescontendingforthechannel.ThusconcatenatingseveralpacketsintoalargesuperpacketcanefcientlyreducetheMAClayercontentionandcollision.However,alongpacketmayneedalongtransmissiontimeduringwhichthechannelqualitymaychangeandhenceencounterahighprobabilityofbiterrors.Thereforeitisnecessarytoconsiderthechannelstatuswhencombiningthepacketstoguaranteethatthetotaltransmissiontimedoesnotexceedthe

PAGE 232

channelcoherencetimeaswellastoconsiderthequeuestatustochecktheavailabilityofpackets.Thisistheproposedadaptivepacketconcatenation(APC)schemeinthischapter.AndtheperformanceofAPCisanalyzedtheoreticallyinbothsinglehopandmultihopadhocnetworks. Therestofthischapterisorganizedasfollows.SectionIIintroducesthebasicsoftheIEEE802.11MACprotocol.TheproposedschemeanditsperformanceanalysisaregiveninSectionIII.Finally,SectionIVconcludesthischapter. ThebasicaccessmethodintheIEEE802.11MACprotocolisDCF(DistributedCoor-dinationFunction),whichisbasedoncarriersensemultipleaccesswithcollisionavoidance(CSMA/CA).Beforestartingatransmission,eachnodeperformsabackoffprocedure,withthebackofftimeruniformlychosenfrom[0,CW-1]intermsoftimeslots,whereCWisthecurrentcontentionwindow.Whenthebackofftimerreacheszero,thenodetransmitsaDATApacket.Ifthereceiversuccessfullyreceivesthepacket,itacknowledgesthepacketbysendinganacknowledgment(ACK).Ifnoacknowledgmentisreceivedwithinaspeci-edperiod,thepacketisconsideredlost;sothetransmitterwilldoublethesizeofCWandchooseanewbackofftimer,andstarttheaboveprocessagain.Whenthetransmissionofapacketfailsforamaximumnumberoftimes,thepacketisdropped.Toavoidcollisionsoflongpackets,theshortRTS/CTS(requesttosend/cleartosend)framescanbeemployed.ThetimingstructureofmessagesequencesareshowninFig. 9 NotethattheIEEE802.11MACalsoincorporatesanoptionalaccessmethodcalledPCF(PointCoordinationFunction),whichisonlyusableininfrastructurenetworkcong-urationsofwirelessLANsanddoesnotsupportmultihopcommunications.Inthischapter,wethusfocusontheIEEE802.11DCF.

PAGE 233

Figure9: RTS/CTSmechanismandbasicaccessmechanismofIEEE802.11 9 .Itconcatenatesseveralpacketsintheinterfacequeuewhichhavethesamenexthopintoasuperpacket.AsuperpacketinsteadoftheoriginalpacketsissenttotheMAClayereachtimewhentheMAClayerisidleandthequeueisnotempty. ThesuperpacketstructureisshowninFig. 9 .Itcontainsoneormoredatapackets.Thesubeldsforeachdatapacketconsistofthreeparts:length,thedatapacketitselfandanoptionalCRCeld.Thelengthsubeldisusedatthereceivertosplitthesuperpacketintotheoriginaldatapackets.TheCRCsubeldisusedtochecktheintegrityofthedatapackettocombatthepossiblechannelbiterrors.Itshouldbeusedifthereceiverenablesselectiveacknowledgementswhichcanindicatewhichdatapacketsarecorruptedbychannelerrorsandneedretransmissions.Ifnotalldatapacketshavenoerrors,thetransmitteronlyneeds

PAGE 234

Figure9: Protocolstack Figure9: Thesuperpacketstructure toretransmitthosecorrupteddatapacketsandreconstructthesuperpacketaccordingtotheavailabledatapacketsinthequeueateachtimeofretransmission. Thelengthlspofasuperpacketisalwayslessthanorequaltoaconcatenationthresh-oldLth.ThisthresholdisdeterminedbythechannelcoherencetimeTccduringwhichthechannelqualityremainsstable[ 113 ].ThetransmissiontimetspofasuperpacketincludesthetransmissiontimeofthephysicalandMAClayeroverheadandthetransmissiontimeforthesuperpacketitself.Andtspmustbelessthanorequaltotcc.Thuswehave forthecasethatthereisnoRTSorCTS,whererdataisthedatarateoftheDATAframe,andTHphyandTHMACarerespectivelythetransmissiontimesofthephysicalandMACheadersofaDATAframe,and

PAGE 235

forthecasethatRTSandCTSareused,whereTRTS,TCTSandTACKarerespectivelythetransmissiontimesoftheRTS,CTSandACKframes. EachtimewhentheMAClayerpicksupDATApacketsfromtheinterfacequeueandstartschannelcontention,APCconcatenatesthepacketattheheadofqueuewithseveralotherpacketsthathavethesamenexthop.Thesepacketsappearinthesuperpacketintheorderthattheyappearinthequeue.TheconcatenationendswhenconcatenatingonemorepacketwillmakethelengthofthesuperpacketexceedLth. Tosupportmultiplechannelrates,APCcalculatesLthusingthecurrenttransmissionrateoftheMAClayer.Therearebasicallytwomethodstodeterminethetransmitraterdata.First,itcanbedeterminedbythehistory.ThetransmitterdeterminesrdataaccordingtothereceivedpowerProfthelastACKframefromthenexthopifthelasttransmissionissuccessful.Otherwiseitusesalowerrateorthelowestavailablerate.Inthesecondmethod,thetransmitraterdataisdeterminedbythereceivedpowerProftheCTSframefromthenexthop.Therstmethoddependsontheresultofprevioustransmissionandmayconcludewithawrongchannelqualitybecauseatransmissionfailurecanresultfromacollisionaswellaspoorchannelquality.ThesecondmethodusestheshortRTS/CTSframestoprobethechannelqualitybeforetheDATAtransmissionandhasamoreaccuratechannelinformation.AlthoughthesecondmethodrequiresRTS/CTSframes,RTS/CTSarealsousefultoshortenthecollisionperiods.Therefore,APCusesthesecondmethodtodeterminerdata. Toutilizemultiplechannelrates,wemustnoticethatdifferentchannelrateshavedif-ferentrequirementsofthereceivedpowerthresholdRXthreshandthesignaltointerferenceplusnoiseratio(SINR).ThewidelyusedIEEE802.11bsupport1,2,5.5,and11Mbps.InEquation( 9.3 ),RXthreshiandCPthreshi(16i64)arethethresholdsrequiredbythehard-waretocorrectlydecodethereceivedsignals.Forexample,therequirementsofaPCMCIASilver/GoldcardbyOrinoccoarethatRXthresh1=94dBm,RXthresh2=91dBm,

PAGE 236

LetRsdenotetheratioofthetimeperiodswithsuccessfultransmissionstothetotaltime.Then,followingthetechniquesinthepapers[ 150 15 ],wehave whereTsistheaveragesuccessfultransmissiontime,Tcistheaveragecollisiontime,isaMAClayeridleslottime,ptisthetransmissionprobabilityofeachnodeinanyslot,nisthetotalnumberofnodesinthenetwork,andpisthecollisionprobabilitythatanodeencounterscollisionwhenevertransmitting.Andfrom[ 68 ], forthecasewheretheRTS/CTSmechanismisused,and timeout+difs;(9.6)

PAGE 237

n1+n(1p)(1(1p)1 n1)Tc(9.10) n1+n(1p)(1(1p)1 n1)Tc(9.11) 15 160 ]forderivationof NowthenetworkthroughputScanbeexpressedasRsmultipliedbytheDATAtrans-missionraterdataexcludingthephysicalandMAClayers'overhead,i.e., Forthesaturatedcasewhereeachnodealwayshasapacketcontendingforthesharedwirelesschannel,Bianchi[ 15 ]derivedtheformulaforthetransmissionprobabilityptatanyslotintermsofp.Consideringaniteretransmissionlimitfollowedbythepacketdropping,wefurtherderivedptinthepaper[ 160 ]as 1p+1+(1p)W(Pi=0(2p)i)2(1p+1) 1p+1+pWPm1i=0(2p)i+W(12mp+1);6m;>m9>=>;(9.9) whereisthemaximumallowedretransmissiontimes,Wistheminimumcontentionwindowsize,and2mWisthemaximumcontentionwindowsize.ByEquations( 9.8 )and( 9.9 ),wecanderivethevalueforp,ptandSforthesaturatedcase,referredas~p,~ptand~S. Forthenon-saturatedcasewherenotallthenodesarecontendingforthechannel,thecollisionprobabilitypissmallerthan~pandhencemayachievealargerthroughput.FromEquation( 9.4 )and( 9.8 ),Scanbeexpressedasthefunctionofp.Sisequalto0,~Sand0

PAGE 238

whenp=0,~pand1,respectively.ToobtainthemaximumvalueofS,denotedbyS,andthecorrespondingvalueofp,denotedbyp,let dpS=0(9.12) Let^pbetherootoftheEquation( 9.12 ).Then IntheAPCscheme,thenetworkthroughputSAPCcanbecalculatedwiththeEquation( 9.11 ),whichisobtainbyexcludingtheAPCoverheadinEquation( 9.10 ).HereLsplistheaveragetotallengthoftheconcatenatedpacketsinasuperpacket.ForthecasethatthepacketlengthLpisxed,wehave wherebLth 9.11 ),TsandTcarecalculatedbyEquations( 9.5 )( 9.6 )and( 9.7 )accordingtotheaveragesuperpacketlengthLsp,whileinEquation( 9.10 ),TsandTcarecalculatedaccordingtotheaveragepacketlengthLp. NowthenetworkthroughputsfortheIEEE802.11protocolwithandwithouttheAPCschemecanbecalculatedbyEquations( 9.10 )and( 9.11 )usingpand~p.ThenumericalresultsareshowninFig. 9 wheren=200.TheparametervaluesoftheIEEE802.11systemareshowninTable 9 FromFig. 9 ,wehavetwoimportantobservations.First,theAPCschemecangreatlyincreasethethroughputwhenthepacketlengthissmallerthanahalfofthecon-catenationthresholdLth.Forthesaturatedcase,thethroughputofAPCschemeisupto3.5timesofthatoftheIEEE802.11protocolwhenthedatapacketlengthisequalto100bytes.AndthemaximumthroughputofAPCschemeisupto2.7timesofthatoftheIEEE802.11

PAGE 239

Figure9: Throughputwhenchannelrateis1Mbps,Lth=2346bytesandRTS/CTSmechanismisused. Table9: IEEE802.11systemparameters ChannelBitRate 1Mbit/s PHYheader 192bits MACheader 224bits LengthofRTS 160bits+PHYheader LengthofCTS 112bits+PHYheader LengthofACK 112bits+PHYheader Initialbackoffwindowsize(W) 32 Maximumbackoffstages(m) 5 Shortretrylimit 7 Longretrylimit 4 protocol.Second,asmallercollisionprobabilityisdesiredtoobtainalargerthroughputsincethemaximumthroughputisalwayslargerthanthesaturatedthroughput.Specically,themaximumthroughputoftheIEEE802.11protocolismuchlargerthanthesaturatedthroughputoftheIEEE802.11protocolespeciallywhenthedatapacketsareshort.Theimprovementrangesfrom4%to32%whenthepacketlengthdecreasesfrom2346to100bytes.WhentheAPCschemeisused,theimprovementrangesfrom4%to7%.Inaddi-tion,asmallercollisionprobabilityisalsorequiredtoachieveashorterdelayandabetterenergyefciency.Itisdesiredtodesignaschemetosupportsmallcollisionprobability

PAGE 240

Figure9: Throughputwhenchannelrateis1,2,5.5and11MbpsandRTS/CTSmecha-nismisused. whileachievingorapproachingthemaximumthroughput.Onesuchschemecanbefoundinthepaper[ 149 ]. Fig. 9 showsthethroughputatdifferentchannelrateswherethepacketlengthLp=512bytesandthechannelcoherencetimeTccisthesameasthatinFig. 9 .IntheAPCscheme,thethroughputapproximatelylinearlyincreasesalongwiththechannelrate.However,thethroughputoftheIEEE802.11protocoldoesnotincreasemuchalongwiththechannelrate.ThisisbecausetherelativeprotocoloverheadismuchlargerforahigherchannelrateintheIEEE802.11protocol.TheimprovementoftheAPCschemeisupto6.2and4.5timeswhenthechannelrateis11Mbpsforsaturatedthroughputandmaximumthroughput,respectively.

PAGE 241

Figure9: Chaintopology. introducesacollisionattheintendedreceiveroftheongoingtransmission.Thiskindofcollisiondependsonthenetworktopologyandisdifculttobecharacterized. Inthissection,wederivethemaximumthroughputthattheIEEE802.11protocolandAPCschemecanachieveinsteadoftheirexactthroughputwhichisdifferentfordifferentnetworkdeployment.Thenwewilldiscusshowtoapproachthismaximumthroughputinawirelessmultihopnetwork.WerststudyamultihopowwhichtravelsthroughachaintopologyasshowninFig. 9 wheresmallcirclesdenotethetransmissionrangeandlargecirclesdenotethecarriersensingrange. Maximumthroughputofamultihopowisachievedwhenthepacketschedulingfullyutilizesthespaceresource,i.e.,schedulingasmanyaspossibleconcurrenttransmissionswithaSINRthatishighenoughforacorrectdecodingatthereceivers.Atanotherhand,nodeswillnotinitiateanynewtransmissionsiftheysenseabusychannelduetothere-quirementofcarriersenseprocedureintheIEEE802.11protocol.Thuswehavetworequirementsformaximumspatialreuse.First,thereisonlyonetransmissioninthecarriersensingrangeofeachnode.Second,thepowerratioofthereceivedsignaltotheinter-ferencesfromothertransmissionsmustbelargerthanorequaltoacertainthresholdasshowninEquation( 9.3 ).Letdenotethepathlossexponent,thenthepowerlevelProfthereceivedsignalequals

PAGE 242

wheredoisthedistancebetweenthetransmitterandareferencepoint,Poisthepowerlevelofthesignalreceivedatthereferencepointanddhisthedistancebetweenthetransmitterandtheintendedreceiver.IntheregularchaintopologyinFig. 9 ,dhisalsothehopdistance. Inthechaintopology,thestrongestinterferencecomesfromtheconcurrenttransmis-sionwhichisclosesttothereceiver.Otherinterferencecanbeneglectedforamuchsmallerpowerlevel.Letdidenotethedistancebetweentwoconcurrenttransmittersinthechaintopology.Forexample,iftransmitter-receiverpair(1,2)and(5,6)canbescheduledtotransmitatthesametime,thendi=4dh.LetPidenotethepowerleveloftheinterferencesignal.GivenacertainrequirementofSINR,wehave ThustheminimumhopdistanceNbetweentwoconcurrenttransmittersequals wheredxeistheceilingfunctionandequalsthelargestintegerlargerthanorequaltox.Thusthemaximumend-to-endthroughputSchainofamultihopowinaregularchaintopologyis whereLp APCisobtainedbyusingasuperpacketinsteadofadatapacket: APC=Lspl whereTsiscalculatedaccordingtothelengthofLsp. Fig. 9 showsthemaximumend-to-endthroughputofamultihopowwithatleastfourhopsintheregularchaintopologywherewesetdo=1m,Po=0dBm,=4and

PAGE 243

Figure9: Maximumend-to-endthroughputofamultihopow. Figure9: Maximumend-to-endthroughputofamultihopow. thedatapacketlengthis512bytes.TherequirementofSINRadoptsthevaluesdiscussedattheendofSection 9.3.1 .Themaximumend-to-endthroughputoftheAPCschemeis1.24,1.53,2.52and4.08timesofthatintheIEEE802.11protocolwhenthechannelrateisequalto1,2,5.5and11Mbps,respectively.Fig. 9 wherethechannelrateis11MbpsshowsthattheAPCschemecanachieveastableandmuchhigherend-to-endthroughputatdifferentpacketlength.ThethroughputoftheAPCschemeis1.62to16.50timesofthatoftheIEEE802.11protocolwhenthepacketlengthdecreasesfrom2246bytesto100bytes. Toachievethemaximumend-to-endthroughput,wemustalleviatethehiddentermi-nalproblemasmuchaspossible.Inthechaintopology,toavoidanodebecomingahiddenterminalandintroducingacollision,thecarriersensingrangemustbelargeenoughtoincludesthenodeswhichcanintroduceenoughinterferencetocorrupttheongoingtrans-mission.Thustheradiusofthecarriersensingrangedcmustsatisfy

PAGE 244

wheretheleftinequationpreventsthecollisionfromthehiddenterminalproblemandtherightinequationmakesitpossibleforthemaximumspatialreuseratio. Besidesthehiddenterminalproblem,wealsoneedtoaddresstheunfairmediumac-cessprobabilityateachforwardingnodetomaximizetheend-to-endthroughput.Onesuchschemecanbefoundinthepaper[ 162 ],whichaddressesbothmediumcontentionandnet-workcongestionandcanwellapproachtheabovemaximumend-to-endthroughput.Foramultihopowinamoregeneraltopology,themaximumend-to-endthroughputdependsonthebottlenecklocationwheretherearethepoorestspatialreuseandthemostinterferencefromotherows.Weleavetheanalysisofsuchtopologytothefuturework.

PAGE 245

Findingapathwithenoughthroughputinmultihopwirelessadhocnetworksisacri-tiquetaskofQoSRouting.Previousstudiesonroutingalgorithmsfocusedonnetworkswithasinglechannelrate.Thecapabilityofsupportingmultiplechannelrates,whichisthoughcommoninwirelesssystems,hasnotbeencarefullystudiedinroutingalgorithms.Inthischapter,werstperformacomprehensivestudyontheimpactofmultiplerates,interferenceandpacketlossratetogetheronthemaximumend-to-endthroughputorpathcapacity.Alinearprogrammingproblemisformulatedtondpathcapacityofanygivenpaths.Thisproblemisalsoextendedtoajointroutingandlinkschedulingoptimizationproblemtondapathwiththelargestpathcapacity.Weprovethatinterferencecliquetransmissiontimeisinverselyproportionaltotheupperboundofthepathcapacity,andhenceweproposeitasanewroutingmetric.Basedontheproposedoptimizationprob-lems,weevaluatethecapabilityofvariousroutingmetricsincludinghopcount,expectedtransmissiontimes,end-to-endtransmissiondelayormediumtime,linkrate,bandwidthdistanceproduct,interferencecliquetransmissiontime,tondapathwithhighthrough-put.Theresultsshowthatinterferencecliquetransmissiontimeisabetterroutingmetricthanothers. 226

PAGE 246

mobileadhocnetworksandwirelessmeshnetworks,wherethereexistsmultihopwirelesscommunication. Tosupportend-to-endcommunicationinthesenetworks,routingalgorithmsplayasignicantroleinndinggoodpathsandforwardingnodesbetweensourcesandtheirdes-tinations.However,ndingagoodpathisnotaneasytaskinawirelessadhocnetworkcomparedinwirednetworksbecausewirelesslinksaresignicantlydifferentfromwiredones.First,wirelesslinksarenotreliableduetochannelerrors.Second,achievablechannelratesmaybedifferentatdifferentlinksbecauselinkqualitydependsondistanceandpathlossexponentbetweentwoneighbors.Third,linksmaynotexistanymorewhenneighborsmoveoutofthecommunicationrange.Fourthbutnotthelast,wirelesstransmissionisbroadcastinnatureandatransmissionoveronelinkwillinterferewithtransmissionsoverotherlinksintheneighborhood. Toaddressthesechallenges,consideringthefeaturesofphysicallayerandMAClayerisamustforagoodroutingalgorithm.However,existingwirelessadhocroutingprotocolstypicallyndrouteswiththeminimumhop-count,shortcomingsofwhichhavebeenrecog-nizedinmultihopwirelessnetworksbymuchpriorresearch.DeCoutoetc.showedinthepaper[ 35 ]thatmanyofshortestpathshavepoorthroughputduetolossratesovertheradiolinksselectedinthesepaths.Theyaccordinglyproposedinthepaper[ 36 ]anewroutingmetricexpectedtransmissioncount(ETX)toconsiderpacketlossratesoverwirelesslinkstoobtainhigherthroughput.Inthepaper[ 75 ],Jainetc.studiedtheimpactofinterferenceonperformanceofmultihopwirelessnetworkwithanNP-completeoptimizationproblem.Theyshowedthatbyconsideringinterference,routesderivedfromtheoptimizationprob-lemoftenyieldnoticeablybetterthroughputthantheshortestpathroutes.Inthepaper[ 77 ]and[ 56 ],JiaandGuptaetc.furtherproposedheuristicalgorithmstoconsiderinter-ferencebysolvinganoptimizationproblemandndpathssatisfyingacertainbandwidthrequirement.

PAGE 247

Besidespacketlossrateandinterference,multiratecapabilityisanothercommonfea-tureofwirelesslinks.Ahigherdataratecanbeusedtoimprovethroughputifabettersignalqualityisobservedoveronelink.However,ahigherdatarateoftenmeansashortertransmissiondistanceandhencemorehopsintheselectedpath.Thedatarateofonelinkisalsosubjecttochangebecauseofatime-varyingchannelandchanginginterferenceintheneighborhood.Noticethatpacketlossratiomaynotbeassignicantasdiscussedinthepaper[ 75 ]ifanauto-rateMACprotocolisadoptedlikeintheIEEE802.11protocol.AlowrateisautomaticallyusedwhenahighpacketlossrateisobservedandhenceleadstoalowpacketlossratebecauseofalessstrictrequirementofSNR(signaltonoiseratio). Notsurprisingly,multiratecapabilityhasagreatimpactonroutingalgorithmsandhencedeservescarefulstudiesinmultihopwirelessadhocnetworks.Itseemsnaturalthatend-to-endthroughputwillbeimprovedifweallowmultipleratestocoexistinthenetwork,whereahigherchannelrateisusedovereachlinkifitcandelivermorepacketsinthesameperiodwiththeconsiderationofpacketlossrates.However,inthepaper[ 84 ],KawadiaandKumarshowedthatasingle-ratewirelessadhocnetworkmayhavebetterperformancethanthenetworkwheremultipleratescoexistiftheshortest-hoproutingalgorithmisused.Thereasonsbehindtheirndingsarethatashortest-hoproutingalgorithmoftenchooselinkswiththelowestchannelratewhileaxedhigherchannelratemaybestillabletogenerateafeasiblepathbetweenthesourceanditsdestinationandleadstoahigherend-to-endthroughput. Severalpapersintheliteraturehavealreadystartedtodesigngoodroutingmetricsinamultiratewirelessadhocnetwork.Inthepaper[ 39 ],Draves,PadhyeandZillproposedtouseweightedcumulativeexpectedtransmissiontime(WCETT)asaroutingmetric.Inthepaper[ 6 ],Awerbuch,HolmerandRubensadoptedmediumtimemetric(MTM).inthepaper[ 155 ],ZhaiandFangstudiedtheimpactofmultirateoncarriersensingrangesand

PAGE 248

spatialreuseratioandaccordinglydemonstratedthatbandwidthdistanceproductandend-to-endtransmissiondelay(samewithmediumtime)arebetterroutingmetricsthanthehopcount. However,thereisstillnocomprehensivestudyonevaluationofthecapabilityoftheseroutingmetricsinmaximizingend-to-endthroughputwithconsiderationofcoexistingmul-tipleratesandtheircloserelationshipwithpacketlossrateandinterference.Thesefactorsmakeitdifculttodesignagoodroutingmetrictondthepathwiththewidestbandwidth.WeuseasimpleexampleinFig. 10 toillustratewhysomeroutingmetricsfailtodoso. InFig. 10 ,allusersareassumedtotransmitoverthesamechannelwithaxedtransmissionpowerandconformtotheIEEE802.11protocols.Supposethehighestachiev-ablechannelrateoverlinksalongpath1fromS1andD1is2Mbps,andthehighestachiev-ablechannelrateoverlinksalongpath2is54Mbps.Apparently,ifSNRrequirementfor1Mbpsislargerthan0dB,transmissionsoveranytwohopsalongpath1cannotbesuccess-fulatthesametime.Thenthemaximumend-to-endthroughputofpath1isproportionalto2 3Mbps.Supposeforthesamereason,thereisalsoonlyonesuccessfultransmissionallowedatatimealongpath2.Themaximumend-to-endthroughputalongpath2is54 12=4:5Mbps.Itissimilarforpath3and4exceptthatpath4passesalargenumberofshorthopsresultinginaverylongend-to-endtransmissiondelay.Supposethattrans-missionsalongpath4canbesimultaneouslysuccessfuleveryother11hopsandsothemaximumend-to-endthroughputofpath4issimilartothatofpath2,i.e.,4:5Mbps.Itisstraightforwardthatpath1willbeselectedfromS1toD1ifaroutingalgorithmminimizesthehopcount.Minimizingtransmissiontimesstillleadstopath1.Minimizingend-to-endtransmissiondelay/mediumtimeormaximizingtheminimumbandwidthdistanceproductoveralllinksalongthepathwillgeneratepath2.Forpath3and4fromS2toD2,hopcount,ETTandend-to-endtransmissiondelayallleadtopath3whilebandwidthdistanceproductleadstopath4withamuchhigherthroughputthanpath3.Itseemsbandwidth

PAGE 249

Figure10: PathsbetweenthesourceSandthedestinationD Inthischapter[ 158 ],weendeavortoaddressallthefactorstogetherinanextendedlinkconictgraphmodel.Alinearprogrammingoptimizationproblemisformulatedtosolvethepathcapacityormaximumend-to-endthroughputofagivenpath.Thesolutionofthepathcapacityinsomescenariosimpliesthatinterferencecliquetransmissiontimeisagoodroutingmetrictondpathswithhighthroughput.Thesolutionoftheoptimizationproblemestablishesafoundationtoevaluatetherelativeperformanceofdifferentroutingmetrics.Themodelisalsoextendedtoajointoptimizationproblemoflinkschedulingandroutingalgorithmtondtheoptimumpathbetweenthesourceandthedestinationthathavethelargestend-to-endthroughput.ThoughthejointoptimizationproblemrequiresacentralizedimplementationandisNP-complete,itprovidesameasurehowgoodtherout-ingmetricsreallyarecomparedtothebestpossibleone.Theresultsshowthatend-to-endtransmissiondelayandinterferencecliquetransmissiontimearethebesttwoamongallthemetricsmentionedaboveonaverage,andinterferencecliquetransmissiontimeconstantlyleadstopathswiththroughputclosetotheoptimumoneandhigherthanthoseobtainedbasedonotherroutingmetrics.Inaddition,interferencecliquetransmissiontimecanndpathswithupto10%morethroughputthanend-to-endtransmissiondelayespeciallywhen

PAGE 250

thedistancebetweenthesourceanditsintendeddestinationislongsayaboutmorethan4hopsintheshortesthoproutingalgorithm.Furthermore,weillustratethatgoodroutingmetricscangeneratepathswithhigherthroughputinamultiratewirelessadhocnetworkthananyroutingmetricsinasingle-ratewirelessadhocnetworkwithanysinglepossiblechannelrate. Therestofthischapterisorganizedasfollows.Section 10.2 discussestheimpactofmultiratecapabilityonthenetworkperformance.Weextendthelinkconictgraphtoconsidermultirate,interferenceandpacketlossratetogethertosolvepathcapacityofanygivenpathinthenetworkinSection 10.3 .InSection 10.4 ,weextendthebellman-fordroutingalgorithmtoutilizeseveraldifferentroutingmetrics.TherelativeperformanceofdifferentroutingmetricsisevaluatedinSection 10.5 .Finally,Section 10.6 concludesthischapter.

PAGE 251

Table10: Signal-to-noiseratioandreceiversensitivity Rates(Mbps) SNR(dB) Receiversensitivity(dBm) 54 24.56 -65 48 24.05 -66 36 18.80 -70 24 17.04 -74 18 10.79 -77 12 9.03 -79 9 7.78 -81 6 6.02 -82 rangeonlyinwhichatransmissioncanbesuccessful.SNRindicateshowmuchinterfer-encecanbetoleratedanddeterminesthespatialreuseratio,i.e.,themaximumnumberofconcurrentsuccessfultransmissionsinacertainarea. Wirelesssystemsnormallysupportmultiplechannelrates,likeUWBand802.11sys-tems.Forexample,alltheIEEE802.11a/b/gstandardssupportmultiplechannelrates.Specically,1,2,5.5,and11Mbpsaresupportedbythe802.11b.6,9,18,24,36,and54Mbpsaresupportedbythe802.11a/g.Differentchannelrateshavedifferentre-quirementsofthereceiversensitivityandSNR.Table 10 showstherequirementofone802.11aproduct[ 143 ].Therefore,transmissionradiusandspatialreuseratiomaybesig-nicantlydifferentfordifferentchannelrates. 32 ]becauseofitshigherrequirementofthereceiversensitivityandSNR.Therefore,usinghigherchannelratesattheforwardingnodesoftenresultsinmorehopsbetweenasourceanditsintendeddestination.Onthecontrary,apathwithsmallestnumberofhopsoftentravelthroughlinkswiththoselowchannelrates,andhencemaysufferfromthroughputloss.

PAGE 252

anytransmission.Theareaaroundonenode,inwhichitcansensetransmissionsfromothernodes,iscalleditscarriersenserange.Therefore,ineachcarriersenserange,thereisatmostonesuccessfultransmitterortransmission. Becauseahigherchannelratehasashortertransmissiondistance,itrequiresmorehopstotravelthroughonecarriersenserangethanalowerchannelrate.Therefore,thespatialreuseratioislowforhighchannelrates.Herethespatialreuseratioismeasuredbythereciprocalofthenumberofhopsbetweenanytwoconcurrentsuccessfultransmissions.Forexample,using54Mbps,themaximumspatialreuseratiomaybeabletoachievedbyschedulingconcurrenttransmissionsatlinksthatareatleast8ormorehopsawayfromeachother[ 155 ].Ontheotherhand,thishopnumberwhenthemaximumspatialreuseratioisachievedcanbe3for1Mbps. TheotherreasonthatahighchannelratehasalowspatialreuseratioisitshighrequirementofSNR.Assumingthatthetransmissionpoweristhesamefortheintendedsignalandtheinterferencesignal,theSNRisproportionalto wheredhishopdistanceorthedistancebetweenthetransmitterandthereceiver,diisthedistancebetweenthereceiverandtheinterferingnode,andisthepathlossexponent.ThusahigherSNRrequiresalargevalueof(di

PAGE 253

Theeffectivedataraterdcanbecomputedas whereTpreambleisthetimenotrelatedwiththechannelraterc,Lplisthelengthofpayloadweintendtotransmit,andLHisthelengthofprotocoloverheadtransmittedwiththechannelraterc.TpreambleincludesthephysicallayerpreambleandmayalsoincludessomeMAClayeroverhead,e.g.interframespacing.LHincludestheMAC,IPandTCPlayers'packethead.Foranexamplein802.11,ifRTS/CTS/ACKaretransmittedwiththebasicrateandDATAistransmittedwiththeselectedchannelraterc,then whereTRTS,TCTS,andTACKarethetimeforthetransmissionofRTS,CTS,andACKframes,respectively.TphyisthetimeforthetransmissionofthephysicalpreambleoftheMACDATAframe.TSIFSandTDIFSaretheinterframespacingtimeofSIFSandDIFS,respectively.IfLplapproachesinnite,rdapproachesrc. GiventhelengthofapacketpayloadLpl,thelargerthechannelrate,thelargerratiothepreambleoccupiesinthetransmissiontimeofapacket,whichmeansaheavierprotocoloverhead.Ahighchannelrateisnormallypreferred,butthecorrespondinglargeprotocoloverheadmustbecarefullyconsidered([ 142 155 ]). 162 151 152 157 ].Theknowledge

PAGE 254

Figure10: Ave-linkchaintopologyanditslinkConictgraph ofpathcapacitycanbeusedtorejectanyexcessivetrafcintheadmissioncontrolforreal-timeservices.Itcanbealsousedinroutingalgorithmstondapathwiththelargestcapacityortoevaluatetheperformanceofdifferentroutingalgorithms.Furthermore,thederivationofpathcapacitymayalsosuggestnovelandefcientroutingmetrics. However,itisnoteasytoderivepathcapacityforpathsinthewirelessadhocnet-works,consideringallthefactorsdiscussedpreviously.Inthissection,werstextendthelinkconictgraphmodeltodescribenecessaryconditionsrequiredbythosefactors.Thenweformulatetheproblemintoalinkschedulingproblemwiththehelpoftheowconictgraph. Inthischapter,weassumethatthereisnopowercontrolschemeorthetransmissionpowerateachnodeisknownbeforelinkscheduling. 10 .Link1and2conictwitheachotherbecausenodeBcannottransmitandreceiveatthesametime.Link1and3conictwitheachotherbecausenodeC'stransmissionwillintroduceenoughinterferenceforthereceivingatnodeB.Link1and4donotconictwitheachotherifnodeD'stransmissionwillnotinterferewiththereceivingatnodeB.

PAGE 255

Thelinkconictgraphcanbeconstructedondifferentphysicallayermodels.Intheprotocolmodel,anyothertransmitterhastobeatleastacertaindistanceawayfromanongoingreceiver.Inthecarriersensingmodel,anyothertransmitterhastobeatleastacertaindistanceawayfromanongoingtransmitter.Inthephysicalmodel,theaggre-gatepowerfromallotherongoingtransmissionsplusthenoisepowermustbelessthanacertainthresholdsothattheSNRrequirementatanongoingreceiverissatised.Inthebi-directionaltransmissionmodel,suchasthe802.11wherethetwo-wayhandshakeDATA/ACKorfour-wayhandshakeRTS/CTS/DATA/ACKareusedforeachtransmission,boththetransmitterandthereceiverofonelinkhastosatisfytherequirementsfromoneormoreoftheabovemodels.Somemixedmodelscanalsobeadopted,suchasamodelconsideringtherequirementsfromboththecarriersensingmodelandthephysicalmodel. Inthischapter,wecallonemodelasadistancemodelifitconsidersthedistancebetweentheconsideredlinkandoneotherlinkatatimelikeinthecarriersensingmodel.Onemodeliscalledasainterferencemodelifitconsiderstheimpactofinterferencepowerlevelfromotherlinkslikeinthephysicalmodel.Amixedmodelconsiderstherequirementsofboththeabovemodels.Allthesemodelscanbecharacterizedbyaweightedconictgraph.Awightwijdescribestheimpactoflinkionlinkj,and wherePrj(i)andPrj(j)arethereceivedpoweratlinkjfromthetransmissionsoverlinkiandj,respectively,PNisthenoisepower,SNRjisrequiredSNRforasuccessfultransmissionatlinkj,andPrj(j) IfPi2S;i6=jwij<1,thetransmissionatlinkjwillbesuccessfulifalllinksbelongingtothesetSaresimultaneouslytransmitting.Ifthisconditionistrueforallj2S,thetransmissionsatallthelinksinScanbescheduledsuccessfullyatthesametime.Such

PAGE 256

asetiscalledanindependentset.IfaddinganyonemorelinkintoanindependentsetSresultsinanon-independentset,Siscalledamaximumindependentset.Forasetoflinks,ifanytwolinksinthesetcannotbescheduledtotransmitsuccessfullyatthesametime,werefertothesetasaninterferenceclique.Ifthesetisnotacliqueanymoreafteraddinganyonemorelink,itisalsoreferredtoasamaximuminterferenceclique. 10.4 )islargerthanorequalto1anddonotconictotherwise,andtheconictrelationshipisindependentonanyotherlinks. Inthissubsection,weassumethatthelinkrateisdeterminedbythereceivedpowerandisequaltothemaximumavailableratesatisfyingtherequirementofreceiversensitiv-ity.WewilldiscussinSection 10.3.4 amoregeneralcasewherethelinkrateisdeterminedbybothreceiversensitivityandsurroundinginterference. LetibetheindexofavailablechannelratesandPse(i)bethereceiversensitivityfortheithchannelrateri.Indexiincreaseswhenchannelrateincreases,andifj>i,rj>riandPse(j)>Pse(i).ThenthelinkratercisdeterminedbythereceivingpowerPratthereceiverofthelink. Givenrcforeachlink,wijcanbecalculatedforanytwolinks,andthelinkconictgraphcanbeconstructedaccordinglyforagiventopology. NowletusdeneanewmetricinterferencecliquetransmissiontimeTCforonecliqueCinthelinkconictgraph,and whereTlisthetransmissiontimeforapacketoverlinkl.ForagivenpathP,ndthesetSofallthemaximuminterferencecliqueCforthelinksbelongingtoP.LetTPbethe

PAGE 257

maximumvalueofTCforallcliquesofPand NoticethatndingallthemaximumcliquesforagraphisaNPhardproblem.However,thenumberoflinksofapathinwirelessnetworksisnormallylimitedtoaverysmallnumber.Thebrute-forcealgorithmcannishndingtheminareasonabletimeifthenumberoflinksissmall. GivenTP,thepathcapacityCPisupperboundedby whereLpisthepacketlength.Thiscanbeeasilyshownbythefollowingfact.TPistheinterferencecliquetransmissiontimeofonecliqueCofP.ConsideringonelinklinCandanyonepacketsuccessfullydeliveredfromthesourcetothedestination,thepackettakestimeTPtotravelthroughallthelinksinC,andlinklcannotscheduleanyothertransmissionduringtheperiodTP.ThatmeansthepackettakesatleasttimeTPatlinkl,andthethroughputatlinklislessthanorequaltoLp Itcanbeshownthatifthereisanoddcycle[ 38 ]inthelinkconictgraph,e.g.inFig. 10 ,theequalsigninEquation( 10.8 )doesnothold.SupposetransmissiontimeofonepacketoveralllinksarethesameandisequaltoT.ItcanbeeasilyshownthatCP=2Lp However,alargeportionofpathsfoundbyroutingalgorithmshavenooddcycleswhenminimizingormaximizingsomemetrics,liketheshortesthoproutingalgorithm.Mostofthesepathsmayhaveauniquefeature:iftwolinksofapathconictwitheachother,allthelinksbetweenthemalongthepathconictwithbothofthem.Wecallthesepathsasdirectroutes,andotherpathsasdetourroutes.Fordirectroutes,theproblem

PAGE 258

Figure10: Apathwithanoddcycleinthelinkconictgraph tondallmaximumcliquecanbesimplied.Tondallthemaximumcliquesincludingonelink,weonlyneedtoconsiderotherlinksclosetothisonealongthepath.Werefertothesecliquesaslocalinterferencecliqueofapath.Fordirectroutes,themaximumvalueofinterferencecliquetransmissiontimeofalllocalcliques,or^TP,isequaltothatforallcliques,orTP.Somepolynomialalgorithmscanbedesignedtondalllocalcliques,whichisomittedinthischapterduetolimitedspace. Fordirectroutes,CP=Lp Itcanbeeasilyshownthattherewillbenoconictinglinksbeingscheduledtotransmitatthesametimesothatitisafeasiblescheduling. Inthissubsection,wedeneanewmetricinterferencecliquetransmissiontimeandshowitcanmoreorlessrepresentthepathcapacity.WewillshowlaterbothmetricsTPand^TP,i.e.,themaximumvalueofinterferencecliquetransmissiontimeofallcliquesandthatofalllocalcliques,canbeusedasroutingmetricstondpathswithhighthroughput,

PAGE 259

and^TPcanbemoreeasilycomputedthanTP.Apparently, Atanytime,atmostoneindependentsetwillbechosentobescheduledtotrans-mitpacketsforalllinksinthatset.Let>0denotethetimesharescheduledtotheindependentsetE,and LetR=fre;(alle2P)gbearowvectorofsizejPj.re=0ife=2E.Otherwise,reistheeffectivedatarateoverlinke,denedinEquation( 10.2 ). Therefore,RisaowvectorthatthenetworkcansupportinthetimesharefortheindependentsetE.WedeneascheduleSasafrequencyvectorS=(:(166M)).Foragivendemandvector~f=ffe,(alle2P)g2RjPj,~fisfeasibleifthereexistsascheduleSsatisfying Pathcapacityisthemaximumend-to-endthroughput,whichonlycountsthetrafctravelingthroughalllinksfromthesourcetothedestination,so

PAGE 260

Now,wecanformatthepathcapacityproblemasfollows. Maximizemine2PfeSubjectto:P166M61P166MR~f=0>0(166M)(10.13) Itcanbeeasilyshownthatthesetofallfeasibledemandvectorsisaconvexset,andgivenafeasibledemandvector~f=ffe,(alle2P)g,thenewvector~f=mine2Pfe(1;1;:::;1)=mine2PfeIisalsofeasible,whereIistheall-onevectorinRjPj.ThustheProblem( 10.13 )canbeconvertedtoalinearprogrammingproblem: MaximizefeSubjectto:P166M61P166MRfeI=0>0(166M);fe>0(10.14) NowwecaninterpretthescheduleSasthefollowinglinkschedulingforagivenpath.Thetimeaxisisdividedintoslotsofduration.Eachtimeslotispartitionedintoasetofsubslotsindexedby(166M),suchthatthethsubslothasalengthofseconds.Inthethsubslot,alllinksinthesetEwillbescheduledtotransmit.Thus,duringeachtimeslotoflength,thethroughputfeoverlinkeis SinceinthesolutionofProblem( 10.14 )feisthesameforalllinks,thepathcapacityisequaltomine2Pfe=fe.

PAGE 261

Inthemulti-interferencemodel,linkconictgraphisaweightedgraphandweightwijbetweenlinkiandjisdenedinEquation( 10.4 ).Independentsetswillbesignicantlydifferentfromthoseobtaininthesingle-interferencemodel,andthehighestachievablelinkrateofeachlinkmaybealsodifferentwhenthelinkisindifferentindependentsetsduetodifferentinterferencelevel. GivenasetoflinksE,theinterferencelevelateachlinkisdeterminedsinceweassumeeachuserusesapredenedtransmissionpower.WhenalllinksinEarescheduledtotransmitatthesametime,SNRatlinkLiinEis wherePriiisthereceivedpowerleveloftheintendedsignalatlinkLi,andPrijforallLj2EnfLigisthereceivedinterferencepoweratlinkLifromthetransmissionatlinkLj.IftwodifferentlinksLiandLjhaveacommonnode,wesetPrji=Prij=1becauseonenodecannottransmitandreceiveatthesametime.Noticethatifbidirectionaltransmissionisadopted,PrijcanbeinterferencelevelofeitherDATAtransmissionorACKtransmission,andwealsoneedtocheckwhethertheSNRrequirementforreceivingbothDATAandACKframesissatisedornotatLinki. IfthereisalinkwhoseSNRislessthantherequirementofthelowestlinkrate,thentransmissionoverthatlinkcannotbescheduledatthesametimewithotherlinksinE,andEisnotanindependentset.Otherwise,Eisanindependentset.Foranindependent

PAGE 262

setE,linkrateofeachlinkinEwillbeselectedasthehighestpossiblechannelratesatisfyingbothrequirementsofreceiversensitivityandSNR. Accordingtotheabovedescriptionofindependentsets,wecanusesomebrute-forealgorithmstondallindependentsetsanddeterminethelinkratesforlinksinthemforonepath.Thenthesamemethodintheprevioussectioncanbeusedtoderivethepathcapacityofanygivenpath. LetP=SiPi,ndallindependentsetsE(166M)andcalculateRforeachEofP.LetI(Pk)isanrowindicatorvectorinRjPj,and Thentheproblemtondthemaximumaggregatethroughputoverallthepathscanbeformattedas MaximizeP16k6KfkSubjectto:P166M61P166MRPkfkI(Pk)=0>0(166M);fk>0(16k6K)(10.18) IfkpathsP1,P2,...,PKbelongtokpairsofsourceanddestination,theproblemformulationisthesameifwewanttomaximizetheaggregatethroughputofallsource-destinationpairs.Iffairnessisconsidered,someotherobjectivefunctions(orconcaveutilityfunctions)canbeused[ 101 ].

PAGE 263

whereDiagf(1pe1);(1pe2);:::;(1pejPj)gisadiagonalmatrixwith(1pei)(16i6jPj)onthediagonal. TheinterferencecliquetransmissiontimeTCbecomesexpectedinterferencecliquetransmissiontimeT0C,and 10.3.2 shouldalsoberecalculatedaccordingly.

PAGE 264

MaximizevSubjectto:Pfj:(i;j)2EgxijPfj:(j;i)2Egxji=8>>>><>>>>:vi=s;0i2Nnfs;tgvi=txij>0;((i;j)2E)P166MR~f=0P166M61;>0(10.21) wherexijistheowfromnodeitonodejoverlinkLij,~fistheowdemandvectorand~f=fxij+xji;((i;j)2E;i
PAGE 265

Inthischapter,wefocusontheunicastandsingle-pathroutingalgorithm.Thereforeweneedtomodifytheaboveproblemintoasingle-pathproblemasfollows: MaximizevSubjectto:Pfj:(i;j)2EgxijPfj:(j;i)2Egxji=8>>>><>>>>:vi=s;0i2Nnfs;tgvi=t06xij6Capijzij;((i;j)2E)Pfj:(i;j)2Egzij61;zij2f0;1gP166MR~f=0P166M61;>0(10.22) whereCapijisthemaximumachievablelinkrateoverlinkLij.Therstthreerowsspecifythatthereisonlyonepathbetweenthesourceandthedestination.Thelinksalongthatpathhavethesameowandallotherlinkshavezeroow.Thisproblemisanmixedinteger-linearproblem. 10.3.2 .Toreducethecomputationtime,localinterferencecliquetransmissiontime(LCTT)canbeused. Ifpacketlossrateisconsidered,theybecomeexpectedtransmissioncount(ETX),expectedend-to-endtransmissiondelay,expectedlinkrate,expectedBDiP,expectedCTT,andexpectedLCTT.Tousetheseroutingmetrics,weshouldndpathstominimizeETX,expectedend-to-endtransmissiondelay,expectedCTT,orexpectedLCTT;ortomaximizeexpectedlinkrateandexpectedBDiP.Thereafter,werefertotheroutingalgorithmsusingthemasmin-hop,min-delay,max-rate,max-BDiP,min-CTTandmin-LCTT,respectively.

PAGE 266

Specically,amin-hoproutingalgorithmndsthepathwiththesmallesthopcountorETX;amin-delayroutingalgorithmndsthepathwiththeshortest(expected)end-to-endtransmissiondelay.Amax-rateormax-BDiProutingalgorithmndsapathwhichhasthewidestbottlenecklink,wherebottlenecklinkofapathisdenedasthelinkwiththelowest(expected)linkrateorsmallestvalueof(expected)BDiPamongallthelinksofthatpath.Amin-CTTormin-LCTTroutingalgorithmndsapathwhichhasthesmallestvalueofbottleneckclique,wherebottleneckcliqueofapathisdenedasthecliquewiththelargestvalueof(expected)CTTorLCTTamongallcliquesorlocalcliquesofthatpath. Amongtheseroutingmetrics,hopcountandend-to-endtransmissiondelayareend-to-endadditiveroutingmetrics.Bellman-Fordalgorithmcanbeusedtominimizethem.Otherroutingmetricscanbeusedwithsomewidestpathroutingalgorithm.Bellman-Fordalgorithmcanbealsousedforthispurposebecauseitiswellsuitedtocomputationofamatrixwiththemaximumbandwidthorlargest/smallestvalueofothermetricsforagivennumberofhops[ 5 ]. Theseroutingmetricscanbealsousedinsomedistributedroutingalgorithms,suchasAODVandDSR.Whenanodeoverhearsarepeatedroutingrequestmessage,itonlyforwardsorrebroadcaststherequestmessagewhentherecalculatedroutingmetricofthepaththatthereceivedrequestmessagetravelsthroughhasabettervaluethanthatforthepreviouslyreceivedrequestmessage,suchasasmallerhopcount.

PAGE 267

396,610m[ 32 ],respectively.Asdiscussedinthepaper[ 155 ],802.11systemhaveverycloseinterferencerangesandoptimumcarriersensingrangesfordifferentchannelrates,soweuseasingleinterferencerange900mforallchannelratesforsimplicity.Thatistosay,aslongastwonodesareatleast900mawayfromeachother,transmissionfromonenodedoesnotinterferewiththereceivingattheother.Thedatapacketsizeis1000bytes.TheIEEE802.11b/gprotocolparametersareadoptedtocalculatetheeffectivedatarateateachlink.1and11Mbpsare802.11bratesand6,18and54are802.11grates.Two-wayhandshakeDATA/ACKisused.BothDATAandACKratesaretransmittedwiththesamelinkrate. Wexthenodenearesttotheupperleftcornerasthesource,andndthepathsfromittoallothernodes.Therefore,therearetotalN1differentsource-destinationpairsorpathsconsideredintheevaluation.Wecomparesevenroutingalgorithmsconsistingofoptimum,min-hop,min-delay,max-band,max-BDP,min-CTTandmin-LCTTroutingalgorithms.Here,optimumrepresentsthemixedinteger-linearproblem( 10.22 )thatfoundthepathwithlargestpathcapacity.Theperformancemetricsarepathcapacity.PathsarecomputedusingtheseroutingalgorithmsandpathcapacityofthesepathsiscomputedbysolvingthelinearprogrammingproblemdenedinEquation( 10.14 ). 10.22 ).NormallyitisaNPhardproblem.Therefore,wecanonlysolvetheproblemforasmalltopologyinareasonabletime.Inthissetofsimulation,25nodesarerandomlydistributedina200mX2500mtopology. Figure 10 showsthepathcapacityofpathsfoundbydifferentroutingalgorithms.Wecanobservethatmin-CTTandmin-LCTTroutingalgorithmscanalwaysndthepathwithapathcapacityequaltotheoptimalvalueinthistopology.Min-delayroutingalgo-rithmscanfoundthepathwithoptimumpathcapacitywhenthesource-destinationdistanceisnotlarge.Howeveritfailstodosowhenthesource-destinationdistanceislargealthough

PAGE 268

Figure10: Pathcapacityfordifferentroutingalgorithms itndsavalueclosetotheoptimumvalue.Max-rateandmax-BDiPmaynotbeabletondapathwiththeoptimumpathcapacitywhetherthesource-destinationdistanceislargeorsmall.Inaddition,min-hoproutingalgorithmhasamuchworseperformancetondapathwithahighthroughputthanallotherroutingalgorithmsbecauseitdoesnotconsiderthemultiratecapabilityofthewirelessnodes. Fig. 10 showspathcapacityofpathsselectedbydifferentroutingalgorithms.First,min-hoproutingalgorithmhasamuchworseperformancethanallotheralgorithms.Sec-ond,min-CTTalwaysndsapathwhichhasalargestpathcapacityamongpathsfoundbyallalgorithms.Third,min-LCTTalmosthasthesameperformanceasmin-CTTforallpairsofsource-destinations.Fourth,min-delayroutingalgorithmcanonlyndapathwithacapacityequaltothatfoundbymin-CTTwhenthesource-destinationdistanceislessthan2000meters,andthepathcapacityis10%lessthanthatfoundbymin-CTTormin-LCTTotherwise.Furthermore,max-rateandmax-BDiProutingalgorithmscannd

PAGE 269

Figure10: Pathcapacityfordifferentroutingalgorithms pathswithcapacityseveraltimesofthatfoundbymin-hopalgorithms,butupto60%lessthanthatfoundbymin-CTTandmin-LCTTroutingalgorithms. Fig. 10 showsthehopcountofpathsfoundbytheseroutingalgorithms.Apparently,min-hoproutingalgorithmndsthepathwithsmallesthopcount.Max-rateandmax-BDiProutingalgorithmsoftenndpathswithaverylargehopcount.Min-delay,min-CTT,andmin-LCTTroutingalgorithmsndpathswithsimilarhopcounts. Fig. 10 showsthesource-destinationdistanceforallthesource-destinationpairs.Thisdistancerangesfromabout0mto3000m.Itismeaningfulwhenconsideredwithothergures.Forexample,whenthesource-destinationdistanceislargerthan2000m,min-hoproutingalgorithmndspathswith4ormorehops,min-delay,min-CTTandmin-LCTTroutingalgorithmsndspathswith7ormorehops,andmin-CTTandmin-LCTTndpathswithcapacitysignicantlylargerthanthatfoundbyotherroutingalgorithms. Fig. 10 showsthesolvingtimeofthepathcapacityproblemdenedinEquation( 10.14 )forallpathsfoundbytheseroutingalgorithms.Sincethisproblemrequiresthe

PAGE 270

Figure10: Pathlengthsfordifferentroutingalgorithms Figure10: Source-destinationdistance

PAGE 271

Figure10: Pathcapacitysolvingtime informationofalltheindependentsets,thesolvingtimealsoincludesthetimetondalltheindependentsetsforallthelinksoftheconsideredpath.Eachpointshowsasolvingtimeforonepath.Wecanobservethatthesolvingtimealmostlinearlyincreaseswiththenumberofhopcountofpaths.Itillustratesthatthepathcapacityproblemcanbesolvedinashorttimewhenthehopcountislessthan22. Table 10 showsthepathndingtimeandthepathcapacitysolvingtimeforalltheroutingalgorithms.Thevaluesinthetableareaggregatevaluesforall399paths.Wecanobservethatmax-CTThasamuchlargervalueofpathndingtimebecausethereisnopolynomialalgorithmtocalculateCTT.Otherroutingalgorithmshaveareasonablepathndingtime.PathcapacitysolvingtimeisapproximatelylineartothehopcountwhichisshowninFig. 10

PAGE 272

Table10: Runtimeofdifferentroutingalgorithms Algorithm Pathndingtime(s) PathCapacitySolvingtime(s) min-hop 1.9840 85.7190 min-delay 10.8280 140.1250 max-rate 4.2030 275.6880 max-BDiP 12.0160 201.6710 min-maxLCTT 24.8750 155.2660 min-maxCTT 289.3130 164.8590 notbethecasewhenhopcountisusedastheroutingmetric[ 84 ].Thetopologyisthesameasthatintheabovesubsection.Min-CTTroutingalgorithmisusedbecauseitcanalwaysndapathwithahigherthroughput.Onlyasinglelinkrate,1,6,11,18,or54Mbps,isallowedinthesingle-ratescenario.Wecomparetheresultsfromsingle-ratescenarioswiththescenariowhereallthesevelinkratesareallowed.Noticethatinthesingle-ratescenarios,ascenariousingalowerlinkratehasmorelinksinthenetworkbecausealowerlinkratehasalargertransmissionrange. Fig. 10 showsthepathcapacityfoundforallthesescenarios.Apparently,muchlargerpathcapacitycanbefoundinthemultiratescenariothanallthesingle-ratescenarios.Noticethat,ifonly54Mbpsisallowedinthenetwork,thenetworkispartitionedintomanypartsandthereisoftennofeasiblepathbetweenasourceanditsdestination.Therefore,pathcapacityisequaltozeroforthescenariowith54Mbpsinthistopology.

PAGE 273

Figure10: Pathcapacityforasingleratenetwork linkrate,end-to-endtransmissiondelay,andbandwidthdistanceproduct,areevaluatedinarandomtopology.Theresultsdemonstratethatinterferencecliquetransmissiontimeisthebestroutingmetrictondapathwithmuchhigherpathcapacitythanotherroutingmetrics.Italsondspathswithpathcapacityequaltotheoptimumonefoundbythejointoptimizationprobleminthesimulatedtopology.

PAGE 274

Inwirelessmultihopadhocnetworks,nodesneedtocontendforthesharedwirelesschannelwiththeirneighbors.Thiscanresultinseverecongestion,packetlossandlongend-to-enddelay,andhenceofferagreatchallengetostreaming,real-time,routingtrafcaswellasTCPtrafc.Differentfromtheproblemsinthewirednetworks,theseproblemsmainlyresultfromthecloseinteractionsbetweenthemediumaccesscontrol(MAC)layerandhigherlayers,andrequireefcientcross-layerdesigns.Inthischapter,wepresentaframeworkofdistributedowcontrolandmediumaccesstomitigatetheMAClayercon-tentions,overcomethecongestion,andincreasetheend-to-endthroughputfortrafcowsacrosssharedchannelenvironments.Thekeyideaisbasedontheobservationthat,intheIEEE802.11MACprotocol,themaximumthroughputforchaintopologyis1=4ofthechannelbandwidthanditsoptimumpacketschedulingistoallowsimultaneoustransmis-sionsatnodeswhicharefourhopsaway.Theproposedfullydistributedschemegeneral-izesthisoptimumschedulingtoanytrafcowwhichmayencounterintra-owcontentionandinter-owcontention.Extensivesimulationsillustratethattheproposedschemewellcontrolscongestionandgreatlyalleviatesmediumcollisions.ItachievesmuchbetterandstablerperformancethantheIEEE802.11MACprotocolintermsofthroughput,delay,fairnessandscalabilitywithlowandstablecontroloverhead. 255

PAGE 275

owswhichpassbytheneighborhood,i.e.,theinter-owcontention,butalsofromthetransmissionsoftheowitselfbecausethetransmissionateachhophastocontendforthechannelwiththeupstreamanddownstreamnodes,i.e.,theintra-owcontention. Thesetwokindsofowcontentionscouldresultinseverecollisionsandcongestion,andsignicantlylimittheperformanceofadhocnetworks.IthasbeenshowninmanypapersthatmultihopadhocnetworksperformpoorlywithTCPtrafcaswellasheavyUDPtrafc([ 91 19 108 140 46 162 147 ]).TheMACprotocolitselfcouldnotsolvethecongestionproblemandoftenaggravatesthecongestionduetothecontentionsinthesharedchannel.FangandMcDonald[ 42 ]studiedhowthethroughputanddelaycanbeaffectedbythepathcoupling,i.e.,theMAClayercontentionbetweennodesdistributedalongnodedisjointpaths,say,inter-owcontention.Theresultsdemonstratedtheneedforthecontrolofcross-layerinteractionsandmethodologiesforcross-layeroptimization. Tothebestofourknowledge,therearenocomprehensivestudyonandgoodsolutionstothecongestioncontrolconsideringtheMAClayercontentionsandthepacketschedulingofmultihoptrafcowsalongtheirselectedpathsinthesharedchannelenvironment.Inthischapter,wepresentaframeworkofnetworklayerowcontrolandMAClayermediumaccesstoaddressthecollisionsandcongestionproblemduetotheintra-owcontentionandinter-owcontention.Basedontheframework,amultihoppacketschedulingalgorithmisincorporatedintotheIEEE802.11DistributedCoordinationFunction(DCF)protocol[ 68 ].Thesalientfeaturehereistogeneralizetheoptimumpacketschedulingforchaintopologytoanytrafcowsingeneraltopology. Theframeworkincludesmultiplemechanisms:fastrelay,backward-pressureconges-tioncontrol,receiver-initiatedtransmissionscheduling,queuespacelimitation,andRoundRobinscheduling.Thefastrelayassignshighpriorityofchannelaccesstothedown-streamnodeswhentheyreceivepackets,whichcanreducealotofintra-owcontentions.Thebackward-pressurecongestioncontrolgivestransmissionopportunitytothecongestednodewhilekeepingitsupstreamnodesfromtransmissions.Thiscouldnotonlyreducealot

PAGE 276

ofcontentionsinthecongestedarea,butalsoquicklyeliminatethecongestion.ItisalsoaquickmethodtonotifythesourcetoslowthesendingratedownbyexploitingtheRTS/CTSoftheIEEE802.11MACprotocol.Thereceiver-initiatedtransmissionschedulingschemeusesathree-wayhandshaketoresumetheblockedowattheupstreamnodeswhenthecongestioniscleared.Itisatimelyandeconomicalapproachwithevenlesscontrolover-headthanthenormalfour-wayhandshaketransmissionintheIEEE802.11protocol.Thequeuespacelimitationforeachowpreventstheirresponsibleapplicationaswellasthecongestedowsfromoccupyingthewholequeuespaceandleavestheresourceforotherresponsibleapplicationsinsteadofthecongestedows.TheRoundRobinschedulingisadoptedinthequeuemanagementtofurtheraddresstheunfairnessproblemduetogreedysources. Thus,altogetherallabovemechanismsprovideaframeworkofdistributedowcon-trolandmediumaccesscontroldesignedtoreducetheMAClayercontentionsandelimi-natethecongestion.Ourcontributionistodevisethesemechanismsforthesharedchannelenvironmentinthemultihopadhocnetworks,andincorporatethemintotheIEEE802.11DCFprotocol.Extensivesimulationstudiesarecarriedouttovalidatetheirperformance.Itturnsoutthatourschemecouldmaintainstableperformancewithhighthroughputin-dependentoftrafcstatus,andimprovetheaggregatedthroughputbyuptomorethan12timesespeciallyforthemultihopowsunderheavytrafcload.Atthesametime,italsoimprovesthefairnessamongowsintermsofend-to-endthroughput,andhasmuchshorterdelayandmuchlowercontroloverheadcomparedtoIEEE802.11DCFprotocol.Moreover,itisscalableforlargenetworkswheretherearemoremultihopowswithlongerpaths. Therestofthischapterisorganizedasfollows.Section 11.2 detailstheimpactofMAClayercontentionsontrafcowsandtheresultingproblems.Section 11.3 introducesourschemeandtheimplementationbasedontheIEEE802.11DCFprotocol.Section 11.4

PAGE 277

evaluatestheperformanceofourschemethroughsimulation.TherelatedworkisdiscussedinSection 11.5 .Finally,weconcludethechapterinSection 11.6 Theintra-owcontentiondiscussedhereistheMAClayercontentionsforthesharedchannelamongnodesofthesameow,whichareineachother'sinterferencerange.Lietal.hasobservedthattheIEEE802.11failstoachievetheoptimumchainscheduling[ 91 ].NodesinachainexperiencedifferentamountofcompetitionsasshowninFig. 111(a) ,wherethesmallcircledenotesanode'svalidtransmissionrange,andthelargecircledenotesanode'sinterferencerange.Thusthetransmissionofnode0ina7-nodechainexperiencesinterferencefromthreesubsequentnodes,whilethetransmissionofnode2isinterferedbyveothernodes.Thismeansthatnode0,i.e.,thesource,couldactuallyinjectmorepacketsintothechainthanwhatthesubsequentnodescanforward.Thesepacketsareeventuallydroppedatthesubsequentnodes.Wecallthisproblemasintra-owcontentionproblem. Inadditiontotheabovecontentionsinsideamulti-hopow,thecontentionsbetweenowscouldalsoseriouslydecreasetheend-to-endthroughput.Iftwoormoreowspassthroughthesameregion,theforwardingnodesofeachowencountercontentionsnotonlyfromitsownowbutalsofromotherows.Thustheprevioushopsoftheseowscouldactuallyinjectmorepacketsintotheregionthanwhatthenodesintheregioncanforward.Thesepacketsareeventuallydroppedbythecongestednodes.AsshowninFig. 11(b)

PAGE 278

(b) Chaintopologyandcrosstopology wheretherearetwoows,oneisfrom0to6andtheotherisfrom7to12.Obviously,node3encountersthemostfrequentcontentionsandhasfewchancestosuccessfullytransmitpacketstoitsdownstreamnodes.Thepacketswillaccumulateatandbedroppedbynode3,9,2,8and1.Wecallthisproblemastheinter-owcontentionproblem. Thesetwoproblemsareverycommonandhaveuniquefeaturesinmultihopadhocnetworks.First,packetforwardingateachhophastocontendforchannelresourcewithothertrafcintheneighborhood.Second,inter-owcontentionnotonlyappearswhenseveralowspassthroughthesameforwardingnode,butalsoexistswhentheows'pathsareclosetoeachothersuchthattheMAClayeronlyallowsonetransmissionatatimetoavoidcollisions.Third,oncethecongestionoccurs,MAClayercontentionsbecomeseveresothattheMAClayerthroughputdecreasesduetotheincreasingcollisionprobability([ 15 154 160 ]).Thisdoesnothelptosolvethecongestionandinsteadresultsinmorepacketsaccumulatinginthequeue. Itiseasytoillustratewhytraditionalcongestioncontrolschemes,suchasTCP,andheavyUDPtrafchavepoorperformanceinadhocnetworksifweconsidertheabovetwoproblems.TCPcannotrespondthecongestionintimeandoftendecreasesthesendingwindowalongtimeafterthecongestionoccurssinceitdependsontheend-to-endfeed-backandtimeoutstoconductcongestioncontrol.Fig. 11 (a)demonstratesthatTCPtrafcintroducesagreatnumberofpacketcollisions.Fig. 11 (b)illustratesmoredetail

PAGE 279

Figure11: TCPperformanceina9-nodechaintopology whythiscouldhappen.Actually,asdiscussedinthepreviousparagraphs,node2,3,and4ina9-nodechainencountermoremediumcontentionthannode0and1,thuspacketsaccumulateatthesenodes,andkeepcontendingforchannelaccess.Thisresultsinseveremediumcollisionandalotofdroppedpackets.TCPacknowledgementsaredelayedandevendroppednotonlyduetotheincreasedMAClayercollisionprobabilitybutalsofortheincreasedqueuelength(noticethateachnodeonlyhasonesharedoutgoinglinkandacorrespondingqueueforalloutgoingpackets).WendinthesimulationsthatTCPsourceoftendetectsthecongestionthroughthesender'stime-outeventsinsteadofduplicatedac-knowledgements.Thisapparentlydegradestheperformanceofcongestioncontrolgreatly.Here,thesimulationsettingsarethesameasthoseofSection 11.4.1 ,anddifferentnumberofTCPowstravelfromnode0tonode8ina9-nodechaintopology.Similarly,sinceUDPtrafchasnocongestioncontrol,itresultsinseverercongestionandintroduceslotsofpacketcollisions,andhencebothend-to-endthroughputanddelaydegradesignicantlywhichwillbeillustratedinlatersimulationresults. ThereforewearguethatagoodsolutiontotheowandcongestioncontrolprobleminadhocnetworksmustconsidertheMAClayercharacteristicsandrespondquicklytothecongestion.Anintuitivesolutiontotheaboveproblemsistoallowthedownstreamnodesandthecongestedonestoobtainthechannelaccesstotransmitpacketswhilekeepingoth-erssilent,andhencesmoothlyforwardeachpackettothedestinationwithoutencountering

PAGE 280

severecollisionsorexcessivedelayattheforwardingnodes.Thismotivatesustodevelopourschemepresentedinthenextsection. 11.3.1Overview OPETincludesfourmajormechanisms.Therstoneistoassignahighpriorityofchannelaccesstothecurrentreceiver.Thiscouldachieveoptimumpacketschedulingforchaintopologyandavoidsevereintra-owcontentionsineachow.Thesecondoneisthehop-by-hopbackward-pressurescheduling.Theforwardingnodesaswellasthesourcearenotiedofthecongestionandthenarerestrainedfromsendingmorepacketstotheirnexthops.ThisefcientlyreducestheMAClayercontentionsduetotheintra-owcontentionandinter-owcontentiononthosecongestednodesbykeepingothernodessilent.Thethirdoneisnottoallowthesourcenodetooccupythewholeoutgoingqueue,whichcouldefcientlypreventtheirresponsibleapplicationsfrominjectingmorepacketsthanthenetworkcouldhandle,andleavemorequeuespaceforotherowspassingthroughthisnode.ThelastoneistheRoundRobinschedulingforthequeuemanagement,whichfurtheralleviatestheunfairnessproblembetweentraversingowsandgreedysourceows.

PAGE 281

itjustreceivesapacket.Thatistosay,thesourcenodetriestoholdthesucceedingpacketsuntiltheprecedingpacketistransmittedoutofitsinterferencerange.Thiscanachieveoptimumschedulingforonewaytrafcintheregularchaintopology. Forexample,inFig. 11(a) ,node1hasthehighestprioritywhenitreceivesonepacketfromnode0andthenforwardsthepackettonode2.Node2immediatelyforwardsthereceivedpacketfromnode1andforwardsittonode3.Itisthesamefornode3,whichimmediatelyforwardsthereceivedpackettonode4.Becausenode0cansensethetrans-missionsofnode1and2,itwillnotinterferewiththesetwonodes.Node0couldnotsendpacketstonode1eitherwhennode3forwardspacketto4becausenode1isintheinterferencerangeofnode3.Whennode4forwardspacketto5,node0couldhavechancetosendpackettonode1.Thesimilarproceduresareadoptedbythesucceedingnodesalongthepath.Node0and4couldsimultaneouslysendpacketstotheirnexthops,andsimilarcasehappenstonodeswhichare4hopsawayfromeachotheralongthepath.Thus,theprocedurecouldutilize1/4ofthechannelbandwidth,themaximumthroughputwhichcanbeachievedbythechaintopology[ 91 ].Foramorerandompath,itispossibleformorethan4hopstointerferewiththersthoptransmission,sothemaximumthroughputislessthan1/4ofthechannelbandwidth.OPET,however,canstillreducelotsofcolli-sionsandapproachthemaximumthroughputthatthetopologycanachievebyallowingthedownstreamnodestoaccessthechannelwithhigherpriority. ToincorporatethisprocedureintotheIEEE802.11DCFprotocol,onesolutionistoassignhigherchannelaccessprioritiestothosepacketswhichhavetraversedmorehops.ItrequiresthattheMAClayersupportsmanydifferentprioritylevels,andneedsthehopcountinformationfromtheroutingprotocol.Currently,weoptforasimplerimplementationwhichsetstheinitialvalueofthebackoffwindowsizeofeachreceiverat8(i.e.,wheneveranodereceivesapacket,itsbackoffwindowissetto8).Whenitnishesthetransmission,theschemeresetsitscontentionwindowsizetothenormalvalue32[ 68 ].TheexampleinFig. 11 showstheoptimumpacketschedulingforthechaintopologyimplementedby

PAGE 282

Figure11: Optimumpacketschedulingforchaintopology ourscheme.Noticethatnode1,2,and3cannotbecomereceiversatthesametimeduetothesharedchannel,andnode0haslowchannelaccesspriorityandcanrarelysendthesucceedingpacketsbeforetheprecedingpacketshavebeenforwardedtonode4. Rule1onlyconsiderstheinterferenceinasingleow.Ifthenexthopofthecurrentreceiverisbusyorinterferedbyothertransmission,thereceivercannotseizethechannelevenwiththehighestpriority.Soweintroducethebackward-pressureschedulingtodealwiththeinter-owcontention. Themechanismincludesatransmissionblockingprocedureandatransmissionre-sumingprocedure.Itrequiresthateachnodemonitorsthenumberofpacketsofindividual

PAGE 283

owinthesharedoutgoingqueue.Letnidenotethenumberofpacketsofowi.Ifnireachesabackward-pressurethreshold,thetransmissionofowifromitsupstreamnodewillbeblocked,andtheupstreamnodeisreferredasarestrictednodeofowiinthefol-lowingdiscussions.Whenthenodesuccessfullyforwardssomepacketstoitsdownstreamnodesothatniislessthanthebackward-pressurethreshold,itinitiatesthetransmissionresumingproceduretoallowtherestrictednodetotransmitpacketsofowi. OurschemeOPETsetsthebackward-pressurethresholdasone,whichindicatestheupperlimitofnumberofpacketsforeachowateachintermediatenode.Thesmallerthevalueis,thelessthemediumcontention.Andoneislargeenoughtobeabletomakefulluseofthechannelbandwidthandissimpletobeimplemented.Noticethatinadhocnetworks,thewirelesschannelissharedbyallthenodesinthesameneighborhood.Atanyonetime,atmostonenodecansuccessfullyaccessthechannelandatmostonepacketcanbesuccessfullytransmittedandreceived.Therefore,atallthenodeswhichareintheinterferencerangeofeachother,ifthetotalnumberofbackloggedpacketsisequaltoorlargerthan1atanytime,thechannelbandwidthwillnotbewastedduetoidleperiod.Forexample,inachaintopologywithmorethan3hops,theoptimumchainthroughputintheIEEE802.11MACprotocolis1/4ofthechainbandwidthandthereforetheoptimumthresholdforthebackward-pressureobjectiveis1/4.Consideringothercontendingtrafcintheneighborhood,thisnumbershouldbesmallertominimizethemediumcontentionaswellastomakefulluseofthechannelbandwidth.Sincefractionalthresholdisdifculttobeimplemented,weoptforthenearestinteger1asthevalueofthisthreshold. ThetransmissionblockingproceduretakesadvantageoftheRTS/CTSexchangeintheIEEE802.11MACprotocoltorestrictthetransmissionfromtheupstreamnodes.AnegativeCTS(NCTS)shouldrespondtheRTSwhentheintendedreceiverhasreachedthebackward-pressurethresholdforthecorrespondingow.Touniquelyidentifyeachow,RTSforthemulti-hopows(RTSM)shouldincludetwomoreeldsthanRTS,i.e.,thesourceaddressandtheowID.RTSforthelasthoptransmissionisnotnecessaryto

PAGE 284

includethesetwoelds,becauseitsintendedreceiveristhedestinationoftheowwhichshouldnotlimititsprevioushopfromsendingpacketstoitself.TheNCTSpackethasthesameformatasCTSexceptthedifferentvalueintheframetypeeld.TheformatofRTSMisshowninFig. 11 Thetransmissionresumingprocedureadoptsthereceiver-initiatedtransmission.Itusesthree-wayhandshakeCTS/DATA/ACKinsteadofthenormalfour-wayhandshakeRTS/CTS/DATA/ACK,becausethedownstreamnodealreadyknowstherestrictednodehaspacketsdestinedtoit.TheCTStoresumethetransmission(CTSR)shouldincludetwomoreeldsthanCTS,thesourceaddressandtheowID,touniquelyspecifytheowasshowninFig. 11 .CTSRaswellasCTShasnoinformationaboutitstransmitterasthatinRTS.Thetwoelds,i.e.,thesourceaddressandtheowID,areusedtouniquelyspecifythenexthopthattheowshouldpassthrough;henceweassigndifferentowIDstotheowsfromthesameapplicationbutwithdifferentpathifmultipathroutingisused.TheprocedureoftransmittingCTSRissimilartothatofRTSandallowsmultipleretransmissionsbeforedroppingit.DifferentmessagesequencesatdifferentsituationsareshowninFig. 11 Thetransmissionresumingprocedurealsoemploysacomplementarymechanism,i.e.,resumingtransmissionbytheupstreamnodeitself.WenoticethatthemobilityinadhocnetworkscouldresultinlinkbreakagefollowedbythetransmissionfailureofCTSR.AndCTSRmayalsobecollidedforseveraltimesandbedropped.Therestrictednodeshouldstartatimer,i.e.,theow-delaytimer,andbeginretransmissionifitsintendedreceiverhasnotsentCTSRbackinalongperiod,whichwesetonesecondinourscheme.Ifthetimeoutvalueistoolarge,theblockedowmaybeblockedforaverylongtimeifCTSRisfailed.Ifitistoosmall,thetransmissionoftheblockedowmayberesumedearlierthanthetimewhenthedownstreamnodeeliminatesthecongestion.Onesecondisatradeoffbetweenthem.

PAGE 285

Inthebackward-pressureschedulingscheme,eachnodeneedstomaintainatable,i.e.,ow-table,torecordtheinformationoftheowswhichcurrentlyhavepacketsintheoutgoingqueue.Atableitemiscreatedwhenaowhastherstpacketintheoutgoingqueue,andwillbedeletedwhenallthepacketsoftheowhavebeenforwardedtothedownstreamnode.Thusthemaximumsizeofthetableisthequeuesizeifallpacketsinthequeuebelongtodifferentowsandthequeueisfull.Theowinformationofeachtableitemincludesthesource-address,ow-ID,number-of-packetsinthequeue,restriction-ag,restriction-start-timeupstream-node-address,andblock-ag.Therestriction-agindicateswhetherthenodeisnotallowedtoforwardthepacketofthisowtothedownstreamnodeandtherestriction-start-timeindicateswhentherestrictionstarts.Theblock-agindicateswhetherthetransmissionoftheupstreamnodeisblocked.Andthealgorithmforbackward-pressureschemeisshowninFig. 11 Figure11: ThepacketformatofRTSMandCTSR OnesimpleexampletoillustratehowourschemeworksisshowninFig. 11(a) andFig. 11(b) .Whencongestionoccursatnode4andnode4couldnotforwardpacket0toitsdownstreamnode5asshowninFig. 11(a) ,theowalongthechainwillaccumulateonepacketateachnodefromnode1tonode4andthenpreventthenodes0,1,2and3fromcontendingforthechanneltoreducethecontentiontothecongestednode4.Aftereliminatingthecongestionatnode4,thetransmissionwillberesumedbythecongestednodeasshowninFig. 11(b) .Noticethatinarandomtopology,thecongestioncanresultfromtheinterferenceorcontentionfromanycrossingand/orneighboringowssuchthat

PAGE 286

RecvNCTS(Packetp)1:SetFlowTable(p,restriction-ag=1,restriction-start-time=now) RecvCTSR(Packetp)1:data=GetDataPktFromQueue(p)2:ifdata!=NULL3:TransmitDATA(data)4:end ResumeTransmissionFromTransmitter(Packetdata)Require:Theretransmissiontimerfortherestrictedowexpiresatthetransmitter.dataisonepacketoftherestrictedowinthequeue1:TransmitRTSM(Packetdata) Thealgorithmsofbackward-pressurescheme theconsiderednodecannotcapturethechannelintime.OPETcanefcientlyforcetheupstreamnodesoftheseowstoyieldthechannelaccessopportunitytothecongestednodeswhichthencanquicklyforwardthebackloggedpacketsandhenceeliminatethecongestion. Itisimportanttonotethatthecontroloverheadofthebackward-pressureschedulingisverysmall.Theinformationofbackward-pressureiscarriedbytheoriginalmessagesequencesRTS/CTSinIEEE802.11.Andtheblockedowisresumedbyathree-way

PAGE 287

Figure11: Messagesequenceforpackettransmission handshakeprocedurewithlessoverheadthantheoriginalfour-wayhandshake.Moreover,ourschemeonlymaintainsseveralsmallentriesforeachactiveow,whichhasatleastonepacketattheconsiderednode.Inamobileadhocnetwork,thenumberofactiveowspernodeisrestrictedbythelimitedbandwidthandprocessingcapability,andhenceisofmuchsmallerorderthaninthewirednetworks,thusthescalabilityproblemshouldnotbeamajorconcerninourscheme. OurschemeOPETsetsthesource-owthresholdasthesmallestintegergreaterthanc+h=4,wherehisthehopcountforeachow.Thequantitycindicatesthemaximumburstofthepacketsthatthequeuecantoleratefortheow.h=4comesfromtheoptimumschedulingofthechaintopologywhichallowssimultaneoustransmissionatnodeswhichare4hopsaway.Consideringthatthechannelissharedbyothertrafcowsinarandomtopology,theachievablethroughputisequaltoorlessthanthatwhenthereisonlyasingleow.Therefore,inageneraltopology,c+h=4islargeenoughtosaturateapathifthesourceisgreedy,andvacatesmorequeuespacefortraversingowsthanthatwhenthereisnosuch

PAGE 288

(b)Thepacketschedulingaftereliminatingthecongestionatnode4 Thepacketschedulingforresolvingcongestion

PAGE 289

sourceself-constraintscheme.ThisthresholdisappliedtoUDPows,andisoptionaltoTCPows.NoticethatTCPcanonlyinjectpacketsuptothereceiver'sadvertisedwindowsizeintothequeue.Furthermore,Chenetal.[ 24 ]havediscoveredintheirsimulationthatTCP'scongestionwindowsizeshouldbelessthankNwhenconsideringtransmissioninterferenceattheMAClayer,where1=8
PAGE 290

RoundRobincannotonlyaddresstheunfairnessproblemduetogreedysource,butalsoprovidefairerschedulingforthetraversingowsthanFCFS(rstcome,rstserved).Ifvariablesizesofpacketsareusedinthenetwork,DecitRoundRobin(DRR)[ 115 ]orSurplusRoundRobin(SRR)[ 3 ]couldbeused.Differentfairqueueingschemeswithintheproposedframeworkwillbeevaluatedinourfuturework. Inoursimulations,thefollowingseveralimportantperformancemetricsareevaluated.

PAGE 291

(b) Simulationresultsforthe9-nodechaintopology(Fig. 11 )andcrosstopol-ogy(Fig. 11(b) ) Inthesimulationstudy,ourschemewillbereferredtoastheOptimumPacketSchedul-ingforEachFlow(OPET),andtheIEEE802.11protocolwithoutthepacketschedulingalgorithmwillbereferredtoastheBasicscheme. 11 ,andthecrosstrafcscenarioshowninFig. 11(b) Fig. 11 showsthatourschemeimprovesthethroughputby55%,120%and33%comparedtotheIEEE802.11inthesethreescenariosunderheavytrafcload,respectively.Weobservethatourschememaintainsasmallandstableend-to-enddelayatalltrafcstatuswhiletheend-to-enddelayincreasesdramaticallywithincreasingtrafcloadintheIEEE802.11protocol.ThereasonisstraightforwardbecausethatourschemereducesalotofMAClayercontentions,i.e.,theintra-owcontentionandtheinter-owcontention,andremovestheexcessivequeueingdelayattheforwardingnodes.

PAGE 292

WeobservefromFig. 11(a) thatwhentheminimumnumberofhopsforeachowincreases,theaggregatedend-to-endthroughputofbothprotocolsdecreases.Thisisreasonablebecausepacketsofmultihopowswithlongerpathhavetopassmorelinksandthusconsumemoreresourceforthesamearrivingtrafc. Fortherandomtrafcwithouthopcountlimitation,ourschemeOPETcouldimprovetheend-to-endthroughputby100%underheavytrafc.ThisisbecausethatOPETreducesalotofchannelcontentionsduetotheintra-owcontentionandinter-owcontention,andtherearemuchlessaccumulatedpacketswhichareeventuallydroppedbytheforwardingnodes.ThereasonthatbasicschemecouldmaintaincertainthroughputunderheavytrafcisthatIEEE802.11MACprotocolgivespreferencetothoseonehoportwo-hopowswhichhavenoormuchlesscontentionsfromhiddenterminals.Theseowscouldcapturethewholebandwidthunderheavytrafcwhichcontributestotheaggregatedend-to-endthroughput.However,otherowswithlongerpathsarestarvedwithzerothroughputsasshowninFig. 11(b) ,whichshowsonerandomexampleofthroughputdistributionamongowsunderheavytrafcandalsoshowstheimprovedfairnessinOPET. Ifsource-destinationpairsofallowsareatleast3hopsaway,OPETcouldstillmaintainhighend-to-endthroughputatheavytrafcloadwhileBasicschemealmostdropstozeroend-to-endthroughput.InBasicscheme,theintra-owcontentioncouldallowthesourcesofmultihopowstoinjectmorepacketsintothenetworkthanthenetworkcanforward.Theinter-owcontentionmakesthesituationworse.ItisnotsurprisingintheBasicschemethatthelongerpaththeowhas,thelowertheend-to-endthroughputitcan

PAGE 293

(b) (c) (d) (e) (f) Simulationresultsfortherandomtopology

PAGE 294

achieve.Byreducingtheintra-owandinter-owcontention,ourschemealwaysmaintainsthehighend-to-endthroughputforallowsatanytrafcloadandtheimprovementismorethan12timesunderheavytrafccomparingtoIEEE802.11protocol. Fig. 11(c) showsthatOPEThasmuchsmallerend-to-enddelaythantheBasicscheme.Also,formultihopows,ourschemeprovidesstableend-to-enddelayinspiteofhightrafcload,whileintheBasicscheme,theend-to-enddelayrapidlyincreaseswiththeofferedload.ThisisbecausethatOPETreducesalotofaccumulatedpacketsintheoutgoingqueueateachnodeandthusgreatlyreducesthequeueingdelay.Inaddition,OPETreducesthecontentionsfromtheintra-owandinter-owcontention,whichcouldalsodecreasethedelayattheMAClayertoaccessthechannel.ItalsoveriesthatinOPETthereisnoseverecongestionwhichcanresultinexcessivequeueingdelayattheforwardednodes. Fig. 11(e) showsthatOPETachievesbettertransmissionefciencyofDATApack-etsashighasabout90%,whiletheBasicschemehasmuchlowervalue,i.e.,evenlessthan5%formultihopows.ThismetricindicatesthattheBasicschemediscardsalotofpacketsthatthesourcessendout,whichhavenotreachedtheintendeddestinations.Thisimpliesthatthesepacketswastealotofwirelessbandwidthandconsumesignicantpower.OPETgreatlyreducesthiskindofwasteandutilizestheresourcetoachievehigherend-to-endthroughput. InOPET,thetransmissionefciencyofDATApacketsisstilllessthan1.Thisisbe-causethatOPETisstillrunningonthecontentionbasedMACprotocol,i.e.,IEEE802.11MACprotocol.ThereexistshiddenterminalproblemwhichresultsinDATApacketcolli-sions. Fig. 11(d) showsthatOPETcouldmaintainsmallandstablenormalizedcontroloverhead.ThisveriesthatOPETcanreducealotofcollisionsattheMAClayerandhencesavealotofunsuccessfulRTS/CTSnegotiationsandDATAtransmissions.TheBasicschemehasmuchhighercontroloverhead,whichrapidlyincreaseswiththeoffered

PAGE 295

loadformultihopows.ThisimpliesthattheBasicschemeisnotappropriateformultihopadhocnetworkswhileOPETisagoodchoiceforthemultihopowsinthesharedwirelesschannelenvironmentandisscalableforlargernetworkswheretherearemoremultihopowswithlongerpaths. Fig. 11(f) showsthatOPETimprovesthefairnessindexbyupto100%comparedtotheBasicscheme.AsintherandomexampleshowninFig. 11(b) ,theBasicschemeonlytakescareofoneortwohopsowswhilestarvingallothermultihopows.Itisnotfairtomultihopowswithlargehopcounts.OPETgivesmuchmorebandwidthtomultihopowswithlargehopcountsthantheBasicscheme.Thefairnessindexisstillmuchlessthanoneinourschemebecausethetrafcdistributionisunbalancedintherandomscenariosandtheowswithshorterpathsstillhaveadvantagesovertheowswithlongerpaths. 108 ].Allresultsareaveragedover30randomsimulationswith300secondssimulatedtimeeach. Thepurposeconsideringthemobilityisonlytoillustratethatourschemecanworkwellinthemobilescenarioswithon-demandroutingscheme.Infact,wendintheex-tensivesimulationsthatmobilitydoesnotchangetheresultsmuch.Therefore,weonlyshowtheaggregatedend-to-endthroughputinFig. 11 ,whichshowsOPEThasabout50%higherthroughputthantheBasicscheme.Allotherperformancemetricsarealsosim-ilarwiththescenariowherethesourceanddestinationarerandomlyselectedwithouthopcountlimitationinthestatictopology. Wealsonoticethatmobilitydecreasesthethroughput.Thisisbecausethattheroutemaybeunavailableduringcertainperiodsduetomobilityalthougheachsourcehasaroute

PAGE 296

Figure11: Simulationresultsfortherandomtopologywithmobility toitsdestinationatthestarttime.Inaddition,theextensivesimulationsalsoindicatethatmobilityincreasestheend-to-enddelaybecausetheroutesearchingandrepairingtimecomesintoplay. 11(a) showsthatourschemeOPETcanreducethepacketcollisionbyabout40%forbothRTSandACKframes.AndthenumberofdroppedTCPpacketsisalsoreducedbyabout80%.Thisveriesthatthehop-by-hopcongestioncontrolcaneffectivelyreducealotofmediumcontentionandcollision.Fig. 11(b) demonstratesthatOPETcanimprovetheaggregatethroughputofTCPowsbyabout5%.AndthefairnessisevenbetterthantheBasicscheme.Fig. 11(c) illustratesthatTCPsourcenodecandetectthecongestionstatusbysimplyobservingitsqueuelengthifOPETisusedandmayaccordinglychangethesendingratetoobtainbetterperformance. NowweexaminetheTCPperformanceinalargernetworkwithgridtopology,whereinter-owcontentionisacommonphenomenon.ThegridtopologyisshowninFig. 11 ,wheretherearetotal100nodes,andone-hopdistanceissetas200meters.16TCPowswith8horizontalonesand8verticalonesrunfor300secondsinthesimulation.Com-paredwiththeBasicscheme,OPETimprovestheend-to-endthroughputfrom547Kbpsto

PAGE 297

(b)Throughputandfairnessinchaintopology (c)QueuelengthforTCPtrafcinchaintopology SimulationresultsfortheTCPtrafc

PAGE 298

Figure11: Gridtopologywith16TCPows 603Kbpsbyabout10%,andreducesthecollidedRTSfrom1015pkt/sto802pkt/sbyabout26%. TheseresultsshowthatOPETcanalsoimproveTCPperformance,althoughTCPowstendtogenerateburstiertrafc.ThisisbecausethatOPETcanreducelotsofpacketdroppingsduetoMACcollisionsaswellasqueueoverowandhenceTCPsourcecon-ductsmuchlessretransmissionsandalsoexperiencesmuchlessoscillationinthesendingwindowsize.TheoptimizationoftheinteractionbetweenTCPandOPETshouldprovidebettersupportforTCPtrafcandwillbestudiedinfuturework.

PAGE 299

RoundRobinschemeintotheframework.Thepreviousoneallocatescertainqueuespacefortraversingpacketstoalleviatethepossibilityofbeingdroppedduetoqueueoverow.Thelatterallowsthatthetraversingowsobtainrelativefairthroughputscomparedwiththeowsgeneratedatthisnodethatmayoftenoccupymostofthequeuespaceandhencemayhavemorechancestobetransmitted.Moreextensivesimulationresultstoillustratetheserelativebenetsareomittedhereduetothepagelimit. 120 48 ]proposedreceiver-initiatedtransmissionschemeswhichworkwellwhentheintendedreceiverknowsexactlythetrafcloadinformation.WangandGarcia-Luna-Aceves[ 132 ]proposedahybridchannelaccessschemewhichcombinesbothsender-initiatedandreceiver-initiatedcollisionavoidancehandshake.Theirschemecouldalleviatethefairnessprobleminsomecaseswithoutsacricingmuchthroughputandsim-plicity,butcannottriggerthedesiredreceiver-initiatedcollisionavoidancehandshakeinsomescenariosduetothelackofowcontentioninformation.Bergeretal.[ 10 ]pre-sentedtwoMAClayerenhancements,i.e.,quick-exchangeandfast-forward,toaddressself-contentioninad-hocnetworks.ThepreviousoneallowsthereceivertoreturnaDATApackettothesenderandthelatterincludesanimplicitRTStothenexthop.Theycouldsavesometransmissionnegotiationprocedures,i.e.,theRTS/CTSexchanges. Inthelastfewyears,severalpapers([ 98 81 ])havebeenreportedforthedistrib-utedpacketschedulingwhichconsiderstheMAClayercollisionsinthemultihopadhocnetworks.Theproposedschemesuseddifferentbackoffwindowsizetoassigndifferentprioritiesforpacketstoaccessthechannel.Luoetal.[ 98 ]constructedtheowcontentiongraphtoachievebetterfairnessamongone-hopowsbetweendifferentnodepairs.Kan-odiaetal.[ 81 ]appliedEDF(EarlyDeadlineFirst)criteriatoobtainsmallerend-to-enddelaythantheoriginalIEEE802.11,althoughcongestionisnotfullyaddressedandthedelaystillincreasesdramaticallywiththeincreasingofferedload.

PAGE 300

Traditionalend-to-endcongestioncontrol,TCP,hasbeenshowntobeinefcientinadhocnetworksinmanyrecentpapers([ 28 29 24 46 50 51 45 140 ]andreferencestherein).MostofthecurrentworktoimproveTCPperformance,suchas[ 23 64 103 95 40 127 46 24 139 ],focusonend-to-endcongestioncontrolmechanismofTCPwithorwithoutnetworklayerfeedback.TheproposedschemesdidnotfullyaddresstheimpactofMAClayerperformanceandstillsufferedfromthesevereMAClayercontentions.Guptaetal.[ 57 ]usedback-pressureconcepttoprovideafairchannelaccesstoTCPowsun-derheavyUDPtrafcwithanimplementationofavirtual,globallyaccessiblearraythatdynamicallyrecordsthequeuelengthsforeachowateachnodeinthenetwork.Monksetal.[ 103 ]conductedsimulationstoillustratethelimitationsofTCP-ELFN[ 64 ]anddis-cussedtheprosandconsofend-to-endcontrolandhop-by-hopcontrol.Theyarguethattheadvantagesofhop-by-hopcontrolmayoutweighitsdrawbacks. Hop-by-hopcongestioncontrolhasbeenstudiedinwirednetworksespeciallyinATMnetworks([ 89 100 ]).ButtheseschemescannotbedirectlyappliedinadhocnetworksduetothecompletelydifferentMACandphysicallayers.Toourbestknowledge,inrecentstudies,only[ 146 ]comprehensivelydiscussedhop-by-hopcongestioncontrolforadhocnetworks.Theauthorsformulatedanoptimizationproblemandstudiedtheend-to-endthroughputunderbothhop-by-hopcongestioncontrolandtheend-to-endcongestioncon-trol.Theirmodelonlyconsideredthechannelsharingforthosenodeswithsameowspassingthrough,anddidnotconsiderothermediumcontentionamongnodeswhichareinthesensingrangeorinterferencerangeofeachother.Comparedtocongestioncontrolschemesforwirednetworks,wecanregardCTSRpacketsinOPETasakindofcreditlikethoseincredit-basedowcontrolscheme[ 89 ].AndNCTSpacketscanberegardedasakindofhop-by-hopsourceQuenchalthoughOPETdoesnotrequirethecooperationoftransportandapplicationlayers[ 111 ]. Tothebestofourknowledge,therearenocomprehensivestudiestoeffectivelyad-dresstheintra-owcontentionandinter-owcontentionproblemsinmultihopmobilead

PAGE 301

hocnetworks,whichresultinseriousproblems,suchasexplosionofcontrolpackets,se-verecollisionsofdatapackets,poorthroughputandfairness,excessivelylongend-to-enddelay,congestion,andpoorscalability.Thus,allthepriorworksonlycontributetotheimprovementofoneortwooftheseperformancemetricswhilesacricingothermetricsmoreorless.Bytacklingthesetwokeyproblemswithanovelcross-layerdesign,ourschemecouldimproveallthesemetricsforbothUDPandTCPtrafc,whichissignicantdeparturefrommostrecentworks.

PAGE 302

ThetraditionalTCPcongestioncontrolmechanismencountersanumberofnewprob-lemsandsuffersapoorperformancewhenappliedinmultihopadhocnetworks.ManyoftheproblemsresultfrommediumcontentionattheMAClayer.Inthischapter,werstillustratethatseveremediumcontentionandcongestionareintimatelycoupled,andthewindowbasedcongestioncontrolalgorithmbecomestoocoarseinitsgranularity,causingthroughputinstabilityandoverlylargedelay.Further,weillustrateTCP'ssevereunfairnessproblemduetothemediumcontentionandthetradeoffbetweenaggregatethroughputandfairness.Then,basedonthenoveluseofchannelbusynessratio,whichweshowisanaccuratesignofthenetworkutilizationandcongestionstatus,anewwirelesscongestioncontrolprotocol(WCCP)hasbeenproposedtoefcientlyandfairlysupportthetransportserviceinmultihopadhocnetworks.InWCCP,eachforwardingnodealongatrafcowexercisestheinter-nodeandintra-nodefairresourceallocationanddeterminesthenetworklayerfeedbackaccordingly.Theend-to-endfeedback,whichisultimatelydeterminedbythebottlenecknodealongtheow,iscarriedbacktothesourcetocontrolitssendingrate.ExtensivesimulationsshowthatWCCPsignicantlyoutperformstraditionalTCPintermsofchannelutilization,delay,andfairness,andeliminatesthestarvationproblem. 283

PAGE 303

Unfortunately,thetraditionalTCPcongestioncontrolmechanismperformsverypoorly,asshowninrecentstudies([ 28 29 24 46 50 51 45 140 ]andreferencetherein).TCPcongestioncontrolhasanimplicitassumption,i.e.,anypacketlossisduetonetworkcon-gestion.However,thisassumptionisnolongervalidintheadhocnetworksaspacketlossesmaywellbeduetochannelbiterrors,mediumcontention,androutefailures. SeveralworkshavepointedoutthatgreedyTCPcanresultinseverecongestioninadhocnetworksandhenceperformancedegradation.Link-RED[ 46 ]wasproposedtomarkordropTCPpacketsaccordingtoobservedpacketcollisions.SubsequentlytheTCPsourcewillreducecongestionwindowsizebeforeitbecomesexcessivelarge.Toavoidcongestion,Chenetal.dynamicallyadjustedthecongestionwindowlimitaccordingtopathlengthofTCPows[ 24 ].Inthepaper[ 139 ],aneighborhoodREDschemewasproposedtoalleviateTCPfairnessproblembyadjustingmarking/droppingprobabilityinlightofobservedchannelinformation. Meanwhile,toalleviatetheadverseimpactofmobility,severalschemeswerepro-posed,suchasthoseinthepapers[ 23 64 103 95 ].Thedesignphilosophyistodistin-guishroutefailuresfromtopologychangesandnetworkcongestionthroughexplicitroutefailurenotications.Otherschemeslike[ 40 127 ],insteadofusingthenetworklayerfeed-back,keeptheTCPstatesunchangedwhenthesourcerstdetectsout-of-orderpacketsandretransmissiontimeout. Inthischapter,wemainlyfocusontheproblemsarisingfrommediumcontention.InSection 12.2 ,weshowthataratebasedcongestioncontrolprotocolismoreappropriatethanitswindowbasedcounterpartinmultihopadhocnetworks.Weillustratetheclosecouplingbetweencongestionandmediumcontention,whichexplainstheinstabilityofTCP.ThenwendthattheoptimumcongestionwindowsizeofTCPmaybelessthanoneeveninverysimpletopology,saychaintopology,inordertomaximizetheend-to-endthroughputandminimizetheend-to-enddelay.

PAGE 304

Nevertheless,itisnotaneasytasktoconductaccurateend-to-endratecontrolintheadhocenvironment,despitethefactthattheexplicitandprecisecongestionfeedbackforend-to-endcontrolhasbeenextensivelystudiedintheInternetandATMnetwork[ 83 2 ].Thisisbecauseeachnodeisindireneedofarobustandeasilymeasuredmetrictoadjustthefeedbackforeachpassingpacket.Whilepacketloss,queuelength,andlinkutilizationaregoodmeasuresforthesewirednetworks,theycannotbedirectlyappliedtoadhocnetworksfortwomainreasons.First,theoccurrenceofpacketlossandlargequeuelengthmayindicatethatseverecongestionhasalreadyhappenedduetomediumcontention,therebyleavingnotimeforthenetworktoreactpromptly.Second,unlikeawiredlinknormallybetweentwonodes,awirelesslinkissharedbyalltheneighboringnodes.Consequently,anychangeinthestatusofthewirelesslinkismuchhardertotrace,whichinturnrendersaccuratecontrolextremelydifcult.Toovercomethisdifculty,weproposetouseanovelmeasuretoreectwirelesslinkstatus.Thechannelbusynessratio,asshowninSection 12.3.1 ,isatimelyandaccuratesignofthenetworkutilizationaswellascongestion. TheninSection 12.3 ,weproposeanewwirelesscongestioncontrolprotocol(WCCP)baseduponthechannelbusynessratio.Inthisprotocol,eachforwardingnodedeterminestheinter-nodeandintra-nodefairchannelresourceallocationandallocatestheresourcetothepassingowsbymonitoringandpossiblyoverwritingthefeedbackeldofthedatapacketsaccordingtoitsmeasuredchannelbusynessratio.Thefeedbackisthencarriedbacktothesourcebythedestination,whichcopiesitfromthedatapackettoitscorrespondingacknowledgement.Finally,thesourceadjuststhesendingrateaccordingly.Clearly,thesendingrateofeachowisdeterminedbythechannelutilizationstatusatthebottlenecknode.Inthisway,WCCPisabletoapproachthemax-minfairness([ 11 ])incertainscenar-ios. WecompareWCCPwithTCPthroughextensivesimulationsinSection 12.4 .WCCPsignicantlyoutperformsTCPintermsofchannelutilization,delay,andfairness.Espe-cially,itsolvesthestarvationproblemsufferedbyTCP.

PAGE 305

Figure12: Chaintopologywith9nodes.Smallcirclesdenotethetransmissionrange,andthelargecirclesdenotethesensingrange Finally,wenotethatWCCPisnotmeanttoattacktheproblemscausedbymobility.Asaresult,WCCPisthemostusefulinstaticmultihopadhocnetworks.However,onecancombinesomeschemesproposedinthepapers[ 23 64 103 95 40 127 ]andWCCPtoalleviatetheperformancedegradationduetomobility.Thiswillbefurtherexploredinourfuturework.ConclusionsaregiveninSection 12.5 12.2.1TCPPerformanceDegradationDuetoCouplingofCongestionandMediumContention 106 ])toconductasetofsimulationsovera9-nodechaintopologyasshowninFig. 12 .OneormoreTCPowswith1000byteslongpayloadtraversefromnode1tonode9.Thepre-computedshortestpathisused,sothereisnoroutingoverhead.Thechannelbandwidthis2Mbps.Simulationseachrunfor300seconds. WecanseefromFig. 12(a) thatTCPtrafcintroducesalotofcollisions.ThoughthereisaretransmissionmechanismforRTSandDATAframesattheMAClayer([ 68 ]),therearestillmanyTCPpacketsdroppedatarateof0:833:63pkts/sbecauseofmediumcontentions.Notethatnopacketlossisobservedforqueueoverow. Fig. 12(b) demonstratesthatTCPtrafcisunstableinthewirelessmultihopenvi-ronment.Theroundtriptime(RT)oscillatesdramatically,andsodoestheinstantaneous

PAGE 306

(b)Flow1'sperformancewhenthereare5ows Simulationresultsfor9-nodechaintopology

PAGE 307

throughput,whichcanbeobtainedbydifferentiatingthenumberofdeliveredpackets(se-quencenumber)withrespecttotime. AlloftheseobservationscanbeattributedtothegreedypropertiesofTCPandthecou-plingofcongestionandcontention.TCPwillcontinuallyincreasethecongestionwindowsizeuntilitdetectsapacketloss.WhenthesendingrateofTCPsourcessurpassthechan-nelcapacity,packetsstarttocumulatealongthepath.Whentheneighboringnodesallhavepacketstotransmit,theykeepcontendingforthechannel.Consequently,morecollisionshappenandhencethechannelcontentiondelayincreases,slowingdowntheforwardingrateandexacerbatingcongestion.Thusthecongestionandcollisionformapositivefeed-backloopuntiltherearesomepacketsdroppedduetocontinualcollisionsandsuchlossesaredetectedbyTCPsourcesfromretransmissiontimeoutsordelayedduplicateACKs. Ifdynamicroutingschemesareusedinmultihopadhocnetworks,thesituationwillbecomeworse.SincetheMAClayercannotdistinguishwhetherthelossesareduetocollisionorunreachablenexthops,itwillreportfalselink/routefailureswhenpacketsaredroppedduetocollisions.Then,theroutinglayerwilllaunchtime-consumingroutesearchorre-routing,therebyincreasingend-to-enddelay. Lietal.[ 91 ]haveshownthat,inchaintopologylikethatinFig. 12 ,themaximumchannelutilizationofachainofadhocnodesis1=4byschedulingthenodesfourhopsawaytotransmitsimultaneously.ThustheoptimalsendingrateRofromthesourcecannotbehigherthanthattomaketheaboveschedulefeasible.Highersendingratewillresultinpacketcollisionsandlosses,andhencelowthroughputandlongdelay.AtrateRo,thepacketisdeliveredtothedestinationinshortesttimewithoutencounteringmuchmediumcollisionandlongqueueingdelay.Also,RTT,denotedbyRTTo,issmall.Assumingthere

PAGE 308

areNTCPowsfromnode1tonode9,theoptimalsendingrateofeachTCPowis whereKeachtcpisthenumberofpacketssentbyeachTCPsourceperRTTo. Accordingtotheaboveoptimalschedule,wecanndouttheoptimalaggregatesend-ingrate.Hereagain,weusesimulationtoillustrateRowhenthe802.11MACisused.CBR/UDPtrafcwiththesamepacketlengthasTCPDATApacketsowsfromnode1tonode9.ReverseCBR/UDPtrafcwiththesamepacketlengthasTCPACKpacketsowsinthereversedirection.TheCBRtrafcintwodirectionshasthesamepacketsendingrate(pkt/s).WegraduallyincreasethesendingrateuntiltheDATApacketdroppingratioduetocollisionislargerthan0:1pkt=sinthe300secondssimulation.Noticethatfurtherin-creasingthesendingrateisfollowedbydramaticincreaseincollisionandpacketdroppingratio.TheresultsaresummarizedinTable 12 ,wheretheperformanceforveTCPowsisalsoincludedforcomparison. Table12: SimulationresultsforTCPandUDPows Trafctype UDP(node1to9) 5TCPows Aggregatethroughput(Kbps) 198 196 Averageend-to-enddelay(s) 0.0695 0.431 RTT(s) Droppedpackets/sduetocollision 0.0931 2.90 Thecorrespondingaggregatesendingrateisabout24:3pkt=sgiventhattheDATApacketdroppingratioduetocollisionislessthan0:1pkt=s.AndRTTo=0:139sandKeachtcp=0:676pkt=RTTwhenN=5.Apparently,themoreTCPowsthereare,thesmallerKeachtcpis.SincetheoptimalsendingrateperRTTislessthanonepacket/RTT,wecanseethatwindowbasedcongestioncontrolprotocolssuchasTCPtendtoovershootthenetworkcapacityastheminimumincreaseinwindowsizeisonepacket.Inother

PAGE 309

words,thegranularityofwindowbasedcongestioncontrolmechanismistoocoarse.Inthissense,windowbasedprotocolsarenotappropriateforsupportingstableandreliabletransportserviceinmultihopadhocnetworks.Therefore,toprovidehighthroughput,shortdelayandstableperformancewithfewpacketcollisions,weoptforanefcientrate-basedcongestioncontrolalgorithmdetailedinSection 12.3 StarvationisasevereunfairnessproblemsufferedbyTCPowsinmultihopadhocnetworks,whichcanbeattributedtothemediumcontention.Thehiddenterminalandreceiverblockingproblems([ 165 161 153 ])arecommoninmultihopadhocnetworks.TogetherwiththegreedinessofTCPows,theycontributetoowstarvationaswellaspacketcollision.Forexample,asshowninFig. 12 ,supposethattherearetwoTCPowspassingthroughthelinks1to2and4to5separately.Whennode4istransmittingpacketstonode5,node2isablockedreceiverofnode1sincenode2sensesthebusychannelandcannotrespondtonode1.Asaresult,node1keepsdoublingitscontentionwindowandretransmittingtheRTSpacketuntildroppingit.Afternode4nishesthetransmission,itresetsitscontentionwindowandhencehashigherprioritythannode1ingrabbingthechannel.Thehiddenterminalproblemmakesnode1'ssituationworse.Supposenode1successfullycontendsforthechannelwithasuccessfulhandshakeofRTSandCTSpacketsandbeginstotransmittheDATApacket.DuringthelongperiodoftheDATAtransmission,node4mayinitiateanewtransmissiontonode5sinceitsensesthechannelisidle.Thistransmissionwillcollidewithnode1'stransmissionatnode2.Thus

PAGE 310

Figure12: Nine-nodechaintopologywithdifferenttrafcdistribution theowpassingthroughthelink1to2willbestarvedifthereisagreedyowpassingthroughthelink4to5. Itisimportanttonotethatthereisatradeoffbetweenfairnessandaggregatethrough-put.Itisknownthatspatialreuseofthechannelbandwidthcanbeachievedbyschedulingsimultaneoustransmissionswhoseregionsarenotinconict.However,assaidabove,differentowsmayexperiencecontentionofdifferentdegree.Achievingfairnessamongthoseowsrequiresallocatingthechanneltoowswithheavecontentionforalongtimeshare,whichcorrespondinglyreducesthechannelreuseandhencetheaggregatethrough-put.Inaddition,whilemaximumthroughputisthechannelbandwidthforaone-hopow,itreducestoonehalf,onethird,andonefourthofthechannelbandwidthforatwo-,three-,andfour-hopow,respectively.Therefore,fairthroughputallocationforowswithdif-ferentpathlengthsinthesameregionhastobeachievedattheexpenseoftheaggregatethroughput. Fig. 12 showsafewmoreexamplesofunfairness.InFig. 12 (b),ow2tra-versesmorehopsandsuffersmoremediumcontention.Consequently,ithastodropmorepacketsandsuffersmoreseriousthroughputdegradationthanow1and3.InFig. 12 (c),ow6suffersnohiddenterminalandreceiverblockingproblemswhileow1does

PAGE 311

andandcouldbestarved.Assigningchannelresourcetoow1willresultinthedecreaseofaggregatethroughputduetothesharedchannelresource.Withperfectscheduling,thethroughputsforthesixowsare(1=12;1=12;1=12;1=12;1=12;1=3)ofthechannelband-widthformax-minfairallocationand(0;1=8;1=8;1=8;1=8;1=2)formaximizingtheag-gregatethroughputandmaintainingfairnessamongow2,3,4,and5atthesametime.Theaggregatethroughputsforthesetwocasesare3=4and1ofthechannelbandwidth,respectively.Clearly,thisdemonstratesthetradeoffbetweentheaggregatethroughputandfairness.WewillgivethesimulationresultsforthesescenariosinSection 12.4 andshowWCCPapproachesthemax-minfairnessincertainscenarios.

PAGE 312

(b)withdifferentpayloadsize Therelationshipbetweenchannelbusynessratioandothermetrics causingseverecongestionandpacketcollision,thecandidateshouldindicatetheavailablebandwidth. Inourpreviouswork[ 150 ],wehaveshownthatchannelbusynessratiomeetsthesetworequirements.AndthemainresultsareshowninFig. 12 .Channelutilizationcuindicatestheratioofthechanneloccupationtimeforsuccessfultransmissionstothetotaltime,normalizedthroughputsindicatestheachievabledatarateforpayloaddividedbythechanneldatarateandisproportionaltocu,andcollisionprobabilityindicatestheaveragepossibilitythateachtransmissionencounterscollision. SeveralimportantresultscanbeobservedfromFig. 12(a) .Firstly,beforechannelutilizationcureachesitspeak,rbisalmostthesameascuandhencecanbeusedtorepresentthenormalizedthroughput.Secondly,afterrbexceedsathresholdwherecureachesitspeak,smallincreaseinrbwillcauseptoincreaseveryfastuntilsaturatedstatusisreached.Thiscaseiscertainlyundesirablesincep,thequeuesize,andthequeuewaitingtimewillallbecomeunacceptablylarge,asindicatedinthepapers[ 160 150 ].Finallyandmostimportantly,theaboveobservationsarealmostindependentofthetotalnumberofnodesintheneighborhood.Thisisaverynicefeaturesincechangesinthenumberofneighborswillnotaffectanode'sperceptionofthechannelutilizationornetworkcongestion,aslongasitreliesontheobservedchannelbusynessratio.

PAGE 313

Nowthatthechannelbusynessratiorbisagoodearlysignofnetworkcongestion,wecanfeedtheobservedrbtotheend-to-endcontrolmechanismtocontrolTCPsourcesandhenceavoidoverloadingthenetwork.Todoso,thekeyistochoosethethreshold,denotedbythb,forrbtoindicatetheinceptionofcongestion.Obviously,thbshouldbechosensuchthat Sincetheperformanceofrbisnotsensitiveton,wecanxnandobservetheeffectofthepayloadsize.Fig. 12(b) showscu,s,andpasafunctionofrb,withdifferentaveragepayloadsizeofDATApacketswhenn=10.Itcanbeobservedthatthesmallertheaveragepayloadsizeis,thesmallerthbshouldbe,usually,i.e.,90%95%.Sincethepayloadsizeof10001500bytesiscommonlyusedinadhocnetworks,wesetthbto92%accordinglyandleave3%spacetoavoidenteringintosaturation. Afterchoosingthb,accordingtoEquation( 12.3 ),wecanestimatetheavailableband-widthofeachnode,denotedbyBWa,asthefollowing: whereBWisthetransmissionrateinbits/sfortheDATApackets,

PAGE 314

aCSMA-basedMACprotocol,workingonthephysicalandvirtualcarriersensingmech-anisms.Thereisalreadyafunctiontodeterminewhetherthechannelisbusyornot,i.e.,thechannelisdeterminedbusywhenthemeasuringnodeistransmitting,receiving,oritsnetworkallocationvector(NAV)([ 68 ])indicatesthechannelisbusy,andisidleotherwise. Inthemultihopadhocnetwork,toovercometheimpactofhiddenterminalandre-ceiverblockingproblems([ 165 161 153 ])ontheestimateofavailablebandwidth,weadoptalittledifferentprocedurefromourpreviouswork([ 150 149 ])todeterminethechannelbusynessratioforwirelessLAN.Specically,thechannelisalsodeterminedbusywhentheMAClayerhasapacketinthebackoffprocedureduetoreceiverblocking.Forexample,whennodeA'sintendedreceiverBisblockedbysomeongoingtransmissionswhichcannotbesensedbyA,i.e.,thechannelresourcearoundBisusedbutthataroundAisidle.WithoutreceivingaresponsefromB,AwilldoubleitsbackoffwindowandkeepsilentforalongertimeduringwhichAsensesthechannelidlebutcouldnotaccommodatemoretrafcsinceBisblocked. 12.4 ),eachnodecouldcalculatethetotalavailablebandwidthforitsneighborhoodbasedonthemeasuredchannelbusynessratioinaperiodcalledaver-agecontrolinterval,denotedby 12.3.5 Todeterminetheavailablebandwidthforeachnode,WCCPaccommodatesthechan-nelresourceSforeachnodeproportionallytoitscurrenttrafcloadSin 12.4 ),wehave Becauseboththeinputtrafcandoutputtrafcofeachnodeconsumethesharedchannelresource,Sshouldincludethetotaltrafc(inbytes),i.e.,thesumofthetotalinputand

PAGE 315

outputtrafc.InFig. 12 (b),forexample,therearethreeowsatnode5,andthetotaltrafcS=r1+r3+2r2,whereri(16i63)isthetrafcofowi. Equation( 12.5 )seemsstraightforward.However,tobetterunderstandit,weneedtoelaborateonit.Therearetwocaseswhenwecomparetheobservedrbwiththb,i.e.,rbthb.Whenrbthb,Sisnegative,meaningwedecreasethetrafc.Inthiscase,however,thelinearrelationshipbetweentheavailablebandwidthandrbnolongerexists,andthecollisionprobabilityincreasesdramaticallyasthetotaltrafcrateincreases.Inaddition,whenthenodeentersintosaturation,bothcollisionprobabilityandrbamounttotheirmaximumvaluesanddonotchangeasthetrafcincreases,althoughthetotalthroughputdecreases.Itthusappearsthatideally,WCCPneedstoaggressivelydecreasethetotaltrafcrate.However,sinceitisdifculttoderiveasimplerelationshipbetweenthetrafcrateandrbwhenrb>thb,WCCPusesthesamelinearfunctionasforthecaserb
PAGE 316

ateachnode.Second,optingoutofaggressivedecreasehelpsachievesmalleroscillationinchannelutilization. WCCPreliesonanAdditive-IncreaseMultiplicative-Decrease(AIMD)policytocon-vergetoefciencyandfairness:IfS>0,allowsincreasethesameamountofthrough-put.AndifS<0,eachowdecreasesthethroughputproportionallytoitscurrentthroughput. BeforedeterminingthepositivefeedbackwhenS>=0,WCCPneedstoestimatethenumberoftheowspassingthroughtheconsiderednode.Again,sincethechannelissharedbybothinputandoutputtrafc,thenumberofowsIusedbyWCCPshouldbedifferentfromtherealnumberofows.Forthoseowsthateitheroriginateorterminateatthenode,thenodecountseachasoneow,whereasforthoseowsthatonlypassthenode,thenodecountseachastwoows,i.e.,oneinandoneout.Forinstance,inFig. 12 (b),I=4fornode2to8,whileI=2fornode1and9.Letrpkdenotethepacketsendingrate(pkt/s)oftheowwhichthekthobservedpacketduringtheperiod whereKisthetotalnumberofdifferentpacketsseenbynodeiin 12 (b),fornode5factork=2forpacketsbelongingtoow2andfactork=1forpacketsbelongingtoow1and3.Ifeachpacketpiggybacksthesourcesendingraterpk,thenodeonlyneedstodothesummationforeachreceivedandtransmitted

PAGE 317

packet,i.e., whereK0isthetotalnumberofreceivedandtransmittedpacketsin Thus,ifS>=0,theincreasingamountoftrafcrateforeachowCp,andperpacketfeedbackpfkwillbe ciI(12.9) IfS<0,WCCPshoulddecreaseeachow'sthroughputproportionallytoitscur-rentthroughput.Alsonoticethattheperpacketfeedbackisinverselyproportionaltotheexpectednumberofpacketsseenbythenodein whereCnisaconstantand ci;(12.12) WCCPaimstomakefulluseofthechannelresourcewhilenotintroducingseveremediumcontention,i.e.,rbshouldbeasclosetothbaspossiblebutneverexceedthbtoomuch.Therefore,whentheaggregatefeedbackofpreviouslypassingpacketsisequaltoS,thenodesetslocalfeedbackvalueaszerountilthenextcontrolintervalstarts.Withthismechanisminplace,thechannelbusynessratiorbshouldbearoundthbatthebottlenecknodesandbesmalleratothernodes. However,itishardtoconvergetoafairresourceallocationsincetheadjustmentbythemultiplicative-decreaselawislimitedifrbiswellcontrolledandalwaysclosetothb.ThusWCCPmanagestotransferresourcefromhighthroughputowstolowthroughput

PAGE 318

owsbyemployingbothincreaselawanddecreaselawwhenjSj<(S+S).LetS+andSdenotetheincreasedanddecreasedtrafcamount,respectively.Also,letpf+kandpfkdenotethepositiveandnegativefeedbackcalculatedbytheincreaselawwithS+andthedecreaselawwithS,respectively.Specically, wheretheadjustmentbythedecreaselawisabout(S+S)ineachcontrolintervalwhenrbisaroundthb.Weset=10%asatradeoffbetweentheconvergencespeedoffairnessandthroughput. 12 .Aleakybucket(per-mitqueue)isattachedtothetransportlayertocontrolthesendingrateofaWCCPsender.Thepermitarrivingraterpoftheleakybucketisdynamicallyadjustedaccordingtotheex-plicitfeedbackfbcarriedinthereturnedACKwheneveranewACKarrives(henceforth,ACKsrefertothetransportlayeracknowledgments).Namely, wherethesettingoffbwillbegivenbelow. Toenablethisfeedbackmechanism,eachWCCPpacketcarriesacongestionheaderincludingthreeelds,i.e.,rp,ci,andfb,whichisusedtocommunicateaow'sstatetotheintermediatenodesandthefeedbackfromtheintermediatenodestothesource.Theeldrpisthesender'scurrentpermitarrivingrate,andtheeldciisthesender'scurrentlyusedcontrolinterval.Theyarelledinbythesenderandnevermodiedintransit.Thelasteld,fb,isinitiatedbythesenderandalltheintermediatenodesalongthepathmaymodifyittodirectlycontrolthepacketsendingrateofthesource.

PAGE 319

Figure12: Ratecontrolmechanism TheWCCPsendermaintainsanestimateofthesmoothedroundtriptimesrttandcalculatesthecontrolintervalcias Whenrpislarge,i.e.,rp>5=srtt,ci=srtt.Otherwise,thisperiodequals5=rp.Thevalueofthecontrolintervalthusensuresthat,onaverage,thereareatleast5datapacketsbeingtransmittedinthisperiod.Iftheperiodistoolong,theadjustmentofthesendingrateissluggishtorespondtotheloadchangealongthepath.Iftheperiodistooshort,theestimationofthefeedbackovershortintervalsatthenodesalongthepathwillleadtoerroneousestimates,andsometimestheremaybenofeedbackreceivedinonecontrolinterval.Thechoiceof5isthetradeoffbetweenthesetwoconsiderations. Initially,whentheWCCPsendersendsouttherstpacketofaow,rp=0,andci=0,indicatingtotheintermediatenodesthatthesenderdoesnotyethaveavalidestimateofthesmoothedroundtriptimesrtt.Thesenderalsoinitializesthefbeldtosuchthatifbandwidthisavailable,thisinitializationallowsthesendertoreachthedesiredrateafteroneci.WhentherstACKreturns,thesendersetsrp=1=rtt,calculatesci,and

PAGE 320

sendsouttheseconddatapacket.Thereafter,aWCCPsendersendsoutadatapacketonlywhenthetransmissionwindowallowsandapermitisavailable. Allthenodesalongtheow'spath,includingtheWCCPsenderandreceiver,keepmonitoringthechannelbusynessratiorb,maintainaper-nodeestimation-controltimerthatissettothemostrecentestimateofaveragecontrolinterval 12.3.3 and 12.3.4 cirpk wherejistheindexforeachpacketobservedin Asfortheoverhead,notethatWCCPdoesnotrequireeachnodetokeepper-owstateinformationandcanscalewelltoanynumberofows.Moreover,thefeedbackcalculationateachintermediatenodeisquitesimple,onlyrequiringafewCPUcyclesforeachpacket. RetransmissiontimerRTOwillexpirewhenthereisapacketloss.Notethatinadhocnetworks,queueoverowrarelyhappensforTCPows.AndpacketlossesmainlyresultfromthefailedtransmissionattemptsattheMAClayerduetocontention,collision,wirelesschannelerror,ormobility-causedroutefailures.Subsequently,thelinkbreakagewillbereportedtotheroutingprotocol,whichmayfurtherdropsubsequentpackets.Noticeinthiscase,theoriginalrouteisbroken,thusthetimeoutsignalsnotonlythepacketloss,butalsotheroutebreakage.Toavoidlongperiodsofpausingandhencewasteofchannel

PAGE 321

capacity,itiswiseforWCCPtosendoutaprobemessageorjustretransmitthelostpacketinperiodicintervalstodetectwhetheranewrouteisestablished. Therefore,theresponseofWCCPtotimeoutisthefollowing.Forthersttimeout,theWCCPsenderretransmitsthecorrespondingpacket,doubletheretransmissiontimer,andresetrpto1=RTO,whereRTOdenotestheretransmissiontimeout.Noteretransmittedpacketshavehigherprioritythannormalpackets.Inotherwords,theretransmittedpacketswillbetransmittedwhenthenextpermitarrives,nomatterwhetherthereareanyotherpacketsinthewindow.Forthesubsequentback-to-backtimeoutsbeforeanewacknowl-edgementsarrives,WCCPdoesnotdoubleitsretransmissiontimeragain,andnordoesitresetrp.ItalsorecordsthetimewhentheretransmissiontimerexpirestodifferentiallytreatthefeedbackinformationcarriedbytheACKsthatarriveafterthetimeoutandrouterepair.ThefeedbackinthoseACKsthatacknowledgethepacketsthataresentpriortothetimeoutissimplyignored,sinceitisverylikelythefeedbackwascalculatedbeforetheroutefailureandhencebecomesoutdated.Bycontrast,thefeedbackinthoseACKsthatacknowledgethepacketsthataresentlaterthanthetimeoutareusedtoadjustthepermitarrivingrateasnormal. Weusenetworksimulatorns2.27toconductthesimulations.Thetransmissionrangeisabout250mandthesensingrangeisabout550m.Wesetthethechannelbandwidthas2Mbpsanduse1000bytesasthepayloadsizeofeachDATApacket. Inthesimulations,werstconsiderthechaintopologyinFig. 12 wherenodesareseparatedby200m,whichissimpleandallowsustoclearlydemonstratetheadvantagesofWCCPoverTCP.Then,weconsiderarandomnetworktopologywithalargenumber

PAGE 322

ofows,inanefforttomodelamorerealisticnetworkenvironment.Thepre-computedshortestpathsareusedunlessotherwiseindicated. Channelutilizationandpacketcollision:WerstconsiderthescenarioinFig. 123 (a).Table 12 showsthatWCCPimprovesthethroughputbyabout8%withonly14%droppedpackets. Table12: PerformanceofWCCPandTCPinchaintopologyofFig. 12 (a) TCP WCCP Throughput(Kb/s) 194.6 209.9 AverageEnd-to-enddelay(s) 0.1844 0.0757 Dropping(pkt/s) 0.834 0.120 InFig. 12(a) ,thechannelbusynessratioispresented.Eachpointinthecurvesisanaveragevalueduringeachsecond.ItcanbeobservedthatWCCPconvergestohighlinkutilizationandstabilizesinanarrowrange,whileTCPfrequentlyoscillatesinalargerange.Infact,thestableandhighchannelutilizationresultsintheimprovementofthroughput.Wealsoobservethatdifferentnodesseedifferentchannelbusynessratio.Sincenode5isinthemiddleofthechainandthusencounterstheheaviestcollisions,itschannelbusynessratioisthelargest.Ontheotherhand,comparedtonode1,node9,asadestination,doesnottransmitanyDATApackets,soitobservesthesmallestchannelbusynessratio. Fig. 12(b) demonstratesthatWCCPhasamuchsmallerroundtriptime,rtt,thanTCP.ForWCCP,theaveragevalueofrttis0:1228s,asopposedto0:2646sforTCP.Fig. 12(c) showsthatWCCPmaintainsamuchsmallerqueuesizeatallthenodesthanTCP.Inaddition,aspointedoutearlier,alargequeuesizekeepsnodebusywithcontendingthechannel,whichincreasescontentionandcausespacketstobedropped.Thus,asmallqueuesizeisdesirable.ThisalsoexplainswhyTCPhasamuchlargerpacketdroppingrate(inpkts/s)thanWCCP.

PAGE 323

(b)Roundtriptime(s) (c)Averagequeuelength Simulationresultsforthenine-nodechaintopologywithoneow

PAGE 324

Figure12: PerformanceofscenarioFig. 12 (b) 12.2.3 .ThesimulationusesthescenariosinFig. 12 (b)and(c). InFig. 12 ,weobservethatTCPcompletelyfailstoguaranteefairnessfortheows.Especially,ow2takesthesmallestshareandow3takesthelargestshareintermsofthroughput.Tosimplifytheexplanation,weonlyconsidertheforwardpathfordatapackets,sincethetransmissionofdatapacketsismuchlongerthanthatofshortACKs.Inthe9-nodechain,nodei+3(16i66)isthehiddenterminalofnodei,becausetheformercannotsensethetransmissionfromthelatterbutwillinterferewiththelatter'sintendedreceiver.Obviously,thesethreeowshavedifferentnumbersofhiddenterminals.Alongthepathofow3,onlynode8isahiddenterminalofnode5,whiletheothertwoows,especiallyow2,suffersevereinterferenceduetomultiplehiddenterminals.Accordingly,thoseowshavedifferentthroughputasshownintheabove,sinceTCPisunabletoensurefairness. Bycontrast,WCCPisabletoallocatefairthroughputtoeachow.Thereasonisthat,bymonitoringthechannelbusynessratioandeachow'strafc,WCCPcanaccuratelycal-culatetheavailablebandwidthofthechannelandfairlyassignittoeachow.Also,sinceWCCPcontrolseachow'sinputtrafcandhencethechannelutilization,itsuccessfully

PAGE 325

Figure12: PerformanceofscenarioFig. 12 (c) reducestheMACcollision.Therefore,wealsodiscoverow2haslessdroppedpacketsthaninthecaseofTCP. WealsosimulatethescenarioinFig. 12 (c).Fig. 12 demonstratesthatTCPfavorsshortows,especiallytheoneortwo-hopows,andpenalizeslongows.Fortheoneortwo-hopows,sinceeachnodealongthepathcansenseothernode'stransmission,thereisnohiddenterminalwithinthepath.Ifthereisnoothercompetingoneortwo-hopowintheneighborhood,theyturnouttoseizeallthebandwidthandobtainhighthroughput.Flow6issuchatwo-hopowandachievesthemaximumthroughputasiftherewerenootherowsintheneighborhood.Asavictim,ow1encountersseverecontentionfromow6andgainsnothroughputatall,althoughthereisapre-computedshortestrouteavailable.Otherfourtwo-hopowscompetewitheachotheralongthesamepathandapproximatelyfairlysharethechannelwithalittlevariation,asseenfromtheirthroughput. WithWCCP,weseethatthestarvingproblemforlongowsisresolved.Flow1achievesalmostthesamethroughputasow2-5,whichsharethesamebottleneck,namely,thenodewiththemaximumnumberofows.Also,ow6takesallthechannelcapacityexceptow1'sshare.Therefore,WCCPapproachesthemax-minfairnessforthisscenarioasdiscussedinSection 12.2.3

PAGE 326

Figure12: Simulationresultsforrandomtopologywithprecomputedpaths:(a)min-imumowthroughputin20runs,(b)minimumowthroughputaveragedover20runs,(c)maximumowthroughputaveragedover20runs,(d)ratioofaveragedmaximumowthroughputtoaveragedminimumowthroughput. 12 conrmsthatthereisatradeoffbetweenthroughputandfairness:fairnessisimprovedattheexpenseoftheaggregatethroughputwhentheoneortwo-hopowsandtheowswithlongerowscoexistinthenetwork.Sincelongowsconsumemoreresourcethanshortowsdowhentransmittingthesameamountoftrafc,ifwegrantalltheowsthesamethroughput,someresourcehastobetakenfromshortowstosupplylongows.Thus,shortowswillsufferthroughputloss.Furthermore,longowsmeanmorehiddenterminalsandmoreMACcollision,andhenceincurmorenodesintotheMACcontention.Anadditionalamountofresourceisthuscon-sumedbythecoordinationofthechannelaccess.Inthisscenario,max-minfairnessisapproachedwithasacriceof1=4ofaggregatethroughputasdiscussedinSection 12.2.3

PAGE 327

(b)AODV Simulationresultsaveragedover20runsintherandomtopology:(1)ag-gregatethroughput(Mbps),(2)fairnessindex,(3)end-to-enddelay(s). Fig. 12 showstherearealwayssomeTCPowshavingbeenstarvedwhileallWCCPowscanobtainacertainamountofthroughput.Theratioofaveragemaximumowthroughputtoaverageminimumowthroughputisdecreasedbyupto1000times.Clearly,WCCPcompletelyeliminatesthestarvationproblem.Furthermore,Fig. 12(a) showsthatWCCPimprovestheJain'sfairnessindexbyabout0.1atapriceof20%45%dropinaggregatethroughput,andtheend-to-enddelayisdecreasedby810times.SimilarresultsarealsoobservediftheondemandroutingprotocolAODVisused,asshowninFig. 12(b)

PAGE 328

theMAClayer,weproposeasystematicsolutionnamedWirelessCongestionControlPro-tocol(WCCP)toaddressthisprobleminbothlayers.Themajorcontributionofthisworkisthree-fold.First,weusesimulationstudiestoshowthatwindow-basedcongestioncon-trolmechanism,saythatofTCP,resultsinpoorandunstableperformanceduetouniquemediumcontentionandhencearguethatrate-basedcongestioncontrolmaybemoreappro-priateforadhocnetworks.Second,weshowthatchannelbusynessratioisagoodsignofnetworkcongestionandavailablebandwidthattheMAClayer,thuscanbeusedasanex-plicitandprecisefeedbackbythetransportlayerinadhocnetworks.Third,weproposeanend-to-endcongestioncontrolprotocol,whichuseschannelbusynessratiotoallocatethesharedresourceandaccordinglyadjuststhesender'sratesothatthechannelcapacitycanbefullyutilizedandfairnessisimproved.WeevaluateWCCPincomparisonwithTCPinvariousscenarios.TheresultsshowthatourschemeoutperformstraditionalTCPintermsofchannelutilization,end-to-enddelay,andfairness,andsolvesthestarvationproblemofTCPows.

PAGE 329

Cross-layerdesignhasbecomeapopularterminthelastfewyearsduetothefactthatthetraditionallayeringnetworkdesignfailsmiserablywhenitisappliedtothewire-lessenvironments,particularlyintheadhocnetworks.Becauseoftheunreliableandunpredictablenatureofthechannels,mostcross-layerdesignfocusesonthephysicallayerandothernetworklayer,whichcanbetypicallyseenaslinkadaptationoropportunis-ticschedulingorotherchannel-awareschemes.Therewasreallynotmuchsignicantprogressinsystematiccross-layerdesignapproacheswhichcanworkeffectivelywhilestillpreservingthedesignsimplicityandscalabilityadvocatedbythelayeringdesignap-proach.Althoughthereweresomeworksoncross-layeroptimization,theyareeithertoocomplicatedtosolveortheresultingsolutionsbecometoosimpletobepractical. Inthisdissertation,weproposeanewcross-layerdesignapproachwhichisbaseduponanin-depthunderstandingofperformanceanddesignissuesofMAClayer.TheMAClayeristheanchoringlayerinourapproachanditreallyconnectsthephysicallayerandthehighernetworkinglayerbecauseitcanreachthephysicallayertogatherlinkinformationwhileaccessinghigherlayerserviceinformation.Aswehaveshown,ourapproachcanwelladdressmediumcontention,QoSprovisioning,fairness,congestioncontrolandroutinginefciencywhilethetraditionallayeringnetworkdesigndoesnotaddresswell.Inthissection,welisttwoimportantissues,fairnessandQoS,inMANETs,whichcanbefurtherstudiedandaddressedfollowingthesameapproachinthiswork. 310

PAGE 330

Figure13: Anoriginaltopologyanditsowcontentiongraph thejointneighborhoodsofthesenderandthereceiver.Andthelevelofcontentionforthewirelesschannelinageographicalregionisdependentonthenumberofcontendingnodesandtrafcstatusintheregion.Second,thereisatradeoffbetweenchannelutiliza-tionandfairness.Spatialreuseofthechannelbandwidthcanbeachievedbyschedulingsimultaneoustransmissionswhoseregionsarenotinconict.However,achievingfairnessrequiresallocatingthechanneltoaowwithalargecontentionforacertaintimeshare,whichcorrespondinglyreducesthechannelreuse.Third,sincethereisnocentralizedcon-trol,nostationisguaranteedtohaveaccurateknowledgeofthecontentioneveninitsownneighborhoodduetothedynamictrafcandtopologyofMANETs.Asaresult,itisverydifculttodesignmechanismstoachievefairness. Manypapers,suchas[ 98 105 67 43 ],begantousetheowcontentiongraphtostudytheowfairnessinMANETs.Fig. 13 showsanoriginaltopologyanditsowcontentiongraph.Therearesixows,eachlyingbetweenapairofneighboringnodes.Clearly,atanytimethereareatmosttwoowsthatcantransmitsimultaneouslywithoutcollidingwitheachother,suchasF1andF4.Translatingthisrestraintintoowcontentiongraph,wecanseethatthereisnolinkbetweenthetwocorrespondingvertexes.Fairnessisachievedbyschedulingthesamechannelresourcetotheowswhichhavethesamelevelofcontentionsinthecontentiongraph,ifpossible. Thetradeoffbetweenfairnessandchannelutilizationcanbedenedasanoptimiza-tionproblem:

PAGE 331

whereNisthenumberofows,xiistherateforowi,fi(xi)isastrictlyconcaveutilityfunction,andwi(>0)istoprovideweightedfairnessorservicedifferentiation.Notethesolutionxiofthisproblemmustcorrespondtoafeasibleschedulingtoachieveit.Theutilityfunctionf(x)canbedenedintermsofowratexas: Itisshownthattheowrateallocationwillapproachthesystem'soptimalfairnessas!0,theproportionalfairnessas!1,andthemax-minfairnessas!1. Sincetheoptimalsolutionoftheaboveproblemdependsonglobaltopology,andisdifculttoachieveinMANETs,severalsub-optimalanddistributedsolutionswerepro-posed.Inthepapers[ 98 67 ],theschemesrequireinformationtobeexchangedbetweenneighborstoconstructalocalowcontentiongraph,andaccordinglycoordinatethechan-nelaccess.Theschemeinthepaper[ 98 ]schedulesadelayinthebackoffprocedureofMAClayeraccordingtotheowdegree.Inthepaper[ 67 ]theminimalcontentionwindowsizeofbackofftimerisdynamicallyadjustedbasedontheobtainedshareofbandwidth.Incontrast,PFCR(ProportionalFairContentionResolution)[ 105 ]andFMAC(FairMAC)[ 43 ]donotneedanyknowledgeofthetopologyofthenetwork.PFCRintroducesaNO CONTENDstateandbeginscontendingforthechannelwithaprobabilityofxiwhenaowhasapackettotransmitandthechannelisidle.Anditobservestheexperi-encedcontentionandaccordinglyadjustsxi.ThebasicideaofFMACistryingtoleteachowtransmitexactlyonepacketinatimeintervaltwhoselengthchangeswiththeloadofthenetworkorthecontentioncontext.Thenumberoftransmissionsinthetimeintervaltservesasthefeedbacksignaltoadjustthecontentionwindoworthetimeinterval. AlltheseschemesachievebetterfairnessthantheIEEE802.11withmoreorlesssacriceofaggregatethroughputincertaintopologies.However,theyarealllimitedtoone-hopows.Thisisbecause,althoughmultihopowsarenotunusualinMANETs,dening

PAGE 332

andachievingfairnessformultihopowsturnsouttobeaverycomplicatedissue.Oneofthereasonsisthatfairnesswithrespecttoend-to-endowrateistightlycoupledwithhigherlayerprotocols,suchasroutingandcongestioncontrol.WehavealreadyproposedadistributedschemetoprovidefairnessinWLANsinChapter 5 .Inthefuture,wewillinvestigateapproachestoimprovethefairnessamongmultihopowsinMANETs. ServicedifferentiationattheMAClayercanbeachievedbyassigningdifferentchan-nelaccessopportunitiestodifferenttypesoftrafc.DifferentbackoffcontentionwindowandDIFSarewidelyusedasdifferentiationtechniquesforsuchpurposes.Forexample,intheEnhancedDistributedCoordinationFunction(EDCF)ofIEEE802.11edraft[ 72 ],trafcisdividedintoeightcategoriesorprioritylevels.Beforetransmitting,eachnodeneedstowaitforthechanneltobeidleforaperiodoftimeassociatedwithitscorrespond-ingtrafccategorycalledArbitrationInterframeSpace(AIFS).Typically,ashorterAIFSandasmallerbackoffcontentionwindowareassociatedwithatrafccategorywithhigherpriority,bywhichEDCFestablishesaprobabilisticprioritymechanismtoallocateband-widthbasedontrafccategories.Inthepaper[ 81 ],similardifferentiationmechanismsarealsoadoptedtoassociateeachpacketwithadifferentpriority,whichisdeterminedbasedonpacketarrivaltimeandpacketdelaybound.Inthisway,delay-sensitivetrafcisbettersupported. Besidesprioritizedchannelaccess,admissioncontrolforthereal-timetrafcisan-otherpowerfultooltosupportbetterQoS.Itcaneffectivelykeepthecongestionofthechannelatalowlevelandreducelongqueueingdelay.Adistributedadmissioncontrol

PAGE 333

algorithm[ 125 ]wasproposedforamulticelltopologywhereeachcellhasabasestation.Bothdataandreal-timetrafcareconsidered.Thisschemereliesontwoalgorithms,i.e.,virtualsource(VS)andvirtualMAC(VMAC),tomeasurethechannelstate.InbothVSandVMACalgorithms,avirtualpacketwasputintheMAClayerorthequeue.Virtualpacketsarescheduledtotransmitontheradiochannelthesamewayasarealpacket,whichmeanschanneltestingandrandombackoffareperformedwhennecessary.Avirtualpacket,however,isnotreallytransmittedwhentheVMACdecidesitwinsthechannel.WhentheestimateddelaybybothVSandVMACexceeds10ms,newreal-timesessionsaredeniedservice.Incontrast,noadmissioncontrolisappliedtodatatrafc.Notethatinadditiontocalladmissioncontrol,real-timetrafcisassignedsmallerbackoffcontentionwindowthandatatrafc.Inthepaper[ 4 ],astatelesswirelessadhocnetworks(SWAN)modelwasproposedforMANETs.SWANuseslocalratecontrolforbest-efforttrafc,andsender-basedadmissioncontrolforreal-timeUDPtrafctodeliverservicedifferentiation. WehavealreadyproposedinChapter 4 acalladmissionandratecontrolschemetoprovidestatisticalQoSguaranteeinwirelessLANs.Inthefuture,wewillinvestigateapproachestosupportbetterQoSthanservicedifferentiationinMANETs.

PAGE 334

[1] I.AdaandC.Castelluccia,DifferentiationmechanismsforIEEE802.11,IEEEINFOCOM'01,Anchorage,Alaska,April2001. [2] Y.Afek,Y.Mansour.Z.Ostfeld,Phantom:asimpleandeffectiveowcontrolscheme,ACMSIGCOMM,Stanford,California,1996. [3] H.Adiseshu,G.Parulkar,andG.Varghese,Areliableandscalablestripingproto-col,Proc.ACMSigcomm,Stanford,California,Aug.1996. [4] G.S.Ahn,A.T.Campbell,A.Veres,andL.H.Sun,Supportingservicedifferen-tiationforreal-timeandbestefforttrafcinstatelesswirelessadhocnetworks(SWAN),IEEETransactionsonMobileComputing,Vol.1,No.3,pp.192-207,2002. [5] G.Apostolopoulos,R.Guerin,S.Kamat,A.Orda,T.Przygienda,andD.Williams,QosroutingmechanismsandOSPFextensions,inRFC2676,InternetEngineer-ingTaskForce,August1999. [6] B.Awerbuch,D.Holmer,andH.Rubens,Themediumtimemetric:Highthrough-putrouteselectioninmulti-rateadhocwirelessnetworks,ToappearintheKluwerMobileNetworksandApplications(MONET)JournalSpecialIssueonInternetWirelessAccess:802.11andBeyond. L.Bao,J.J.Garcia-Luna-Aceves,Distributedchannelaccessschedulingforadhocnetworks,SubmittedforpublicationinIEEETransactionsonNetworking. [8] Y.Bejerano,S.-J.Han,,andL.Li,FairnessandloadbalancinginwirelessLANsusingassociationcontrol,inACMMobiCom,Philadelphia,Pennsylvania,USA,Sept.2004. [9] Y.BejeranoandR.Bhatia,MiFi:aframeworkforfairnessandQoSassuranceincurrentIEEE802.11networkswithmultipleaccesspoints,InProc.IEEEINFO-COM,HongKong,2004. [10] D.Berger,Z.Ye,P.Sinha,S.V.Krishnamurthy,M.Faloutsos,S.K.Tripathi,TCPfriendlymediumaccesscontrolforad-hocwirelessnetworks:alleviatingselfcon-tention,Proc.IEEEMASS,FortLauderdale,Florida,USA,Oct.2004. [11] D.BertsekasandR.Gallager,Datanetworks,SecondEdition,PrenticeHall,Engle-woodCliffs,NJ,1992. 315

PAGE 335

[12] P.Bhagwat,P.Bhattacharya,A.Krishna,andS.Tripathi,EnhancingthroughputoverwirelssLANsusingchannelstatedependentpacketscheduling,IEEEINFO-COM,SanFrancisco,California,USA,March1996. [13] V.Bharghavan,Performanceevaluationofalgorithmsforwirelessmediumaccess,Proc.IEEEInternationalComputerPerformanceandDependabilitySymposium,Durham,NorthCarolina,1998. [14] V.Bharghavan,A.Demers,S.Shenker,andL.Zhang,MACAW:amediaaccessprotocolforwirelessLAN's,Proc.ofACMSIGCOMM,London,UK,1994. [15] G.Bianchi,PerformanceanalysisoftheIEEE802.11distributedcoordinationfunc-tion,IEEEJ.Sel.AreasCommun.,vol.18,pp.535-547,Mar.2000. [16] G.Bianchi,L.Fratta,andM.Oliver,PerformanceevaluationandenhancementoftheCSMA/CAMACprotocolfor802.11wirelessLANs,InProc.IEEEPIMRC,Taipei,Taiwan,1996. [17] G.BianchiandI.Tinnirello,KalmanlterestimationofthenumberofcompetingterminalsinanIEEE802.11network,inIEEEINFOCOM,SanFranciso,CA,USA,2003. [18] S.Blake,D.Black,M.Carlson,E.Davies,Z.Wang,andW.Weiss,RFC2475Anarchitecturefordifferentiatedservices,InternetEngineeringTaskForce,1998. [19] J.Broch,D.A.Maltz,D.B.Johnson,Y.Hu,andJ.Jetcheva,Aperformancecom-parisonofmultihopwirelessadhocnetworkroutingprotocols,Proc.ACM/IEEEMobiCom,Dallas,Texas,USA,Oct.1998. [20] F.Cali,M.Conti,andE.Gregori,IEEE802.11protocol:designandperformanceevaluationofanadaptivebackoffmechanism,IEEEJSAC,vol.18,pp.1774-1786,Sept.2000. [21] F.Cali,M.Conti,andE.Gregori,TuningoftheIEEE802.11protocoltoachieveatheoreticalthroughputlimit,IEEE/ACMTrans.onNetworking,vol.8,pp.785-799,Dec.2000 [22] F.Cali,M.Conti,andE.Gregori,IEEE802.11wirelessLAN:capacityanalysisandprotocolenhancement,Proc.IEEEINFOCOM,SanFrancisco,CA,USA,March,1998 [23] K.Chandran,S.Raghunathan,S.Venkatesan,andR.Prakash,Afeedback-basedschemeforimprovingTCPperformanceinadhocwirelessnetworks,IEEEPer-sonalcommunications,vol.8,pp.34-39,Feb.2001. [24] K.Chen,Y.Xue,andK.Nahrstedt,OnsettingTCP'scongestionwindowlimitinmobileadhocnetworks,Proc.IEEEICC,Anchorage,Alaska,May2003.

PAGE 336

[25] X.Chen,W.Liu,H.Zhai,andY.Fang,Location-awareresourcemanagementinmobileadhocnetworks,toappearinACMWirelessNetworks,2004. [26] X.Chen,H.Zhai,andY.Fang,EnhancingtheIEEE802.11einQoSsupport:analy-sisandmechanisms,inProc,ofTheSecondInternationalConferenceonQualityofServiceinHeterogeneousWired/WirelessNetworks(QShine05),Orlando,Florida,USA,Aug.2005. [27] X.Chen,H.Zhai,X.Tian,andY.Fang,SupportingQoSinIEEE802.11ewirelessLANs,toappearinIEEETransactionsonWirelessCommunications,2005. [28] X.Chen,H.Zhai,J.Wang,andY.Fang,TCPperformanceovermobileadhocnet-works,CanadianJournalofElectricalandComputerEngineering(CJECE)(Spe-cialIssueonAdvancesinWirelessCommunicationsandNetworking),Vol.29,No.1/2,p129-134,January/April2004. [29] ,AsurveyonimprovingTCPperformanceoverwirelessnetworks,ResourceManagementinWirelessNetworking,editedbyM.Cardei,I.CardeiandD.-Z.Du,KluwerAcademicPublishers,2004. [30] D.W.Choi,Framealignmentinadigitalcarriersystem-atutorial,IEEECommu-nicationsMagazine,Volume:282,Feb.1990 [31] S.Choi,J.Prado,S.Mangold,andS.Shankar,IEEE802.11econtention-basedchannelaccess(EDCF)performanceevaluation,Proc.IEEEICC,Anchorage,Alaska,May,2003. [32] [33] D.Clark,S.Shenker,andL.Zhang,Supportingreal-timeapplicationinanin-tegratedservicespacketnetwork:architectureandmechanism,Proc.ACMSIG-COMM,Baltimore,Maryland,USA,1992. [34] C.Coutras,S.Gupta,andN.B.Shroff,Schedulingofreal-timetrafcinieee802.11wirelesslans,WirelessNetworks,6(6):457-466,2000. [35] D.S.J.DeCouto,D.Aguayo,B.A.Chambers,andR.Morris,Performanceofmultihopwirelessnetworks:Shortestpathisnotenough,inProc.theFirstWork-shoponHotTopicsinNetworks(HotNets-I),Princeton,NewJersey,USA,October2002. [36] D.S.J.DeCouto,D.Aguayo,J.Bicket,andR.Morris,Ahigh-throughputpathmetricformulti-hopwirelessrouting,inProc.ACMMobicom,SanDiego,CA,USA,September2003. [37] J.Deng,B.Liang,andP.K.Varshney,TuningthecarriersensingrangeofIEEE802.11MAC,inProc.IEEEGLOBECOM,Dallas,Texas,USA,Dec.2004.

PAGE 337

[38] R.Diestel,GraphTheory,3rded.Springer,NewYork,2006. [39] R.Draves,J.Padhye,andB.Zill,Routinginmulti-radio,multi-hopwirelessmeshnetworks,inProc.ACMMobicom,Philadelphia,PA,USA,September2004. [40] T.D.DyerandR.V.Boppana,AcomparisonofTCPperformanceoverthreerout-ingprotocolsformobileadhocnetworks,Proc.ACMMobihoc,LongBeach,Cali-fornia,USA,Oct.2001. [41] K.FallandK.Varadhan,editors,NSnotesanddocumentation,TheVINTProject,UCBerkeley,LBL,USC/ISI,andXeroxPARC,April,2002 [42] Y.FangandA.B.McDonald,Cross-layerperformanceeffectsofpathcouplinginwirelessadhocnetworks:powerandthroughputimplicationsofIEEE802.11MAC,Proc.IEEEIPCCC,Phoenix,Arizona,USA,Apr.2002 [43] Z.FangandB.Bensaou,Fairbandwidthsharingalgorithmsbasedongametheoryframeworksforwirelessad-hocnetworks,Proc.IEEEInfocom,HongKong,China,March2004. [44] C.H.FohandM.Zukerman,PerformanceanalysisoftheIEEE802.11MACpro-tocol,Proc.EuropeanWireless,Florence,Italy,Feb.2002. [45] Z.Fu,X.Meng,andS.Lu,HowbadTCPcanperforminmobilead-hocnetworks,IEEESymposiumonComputersandCommunications,Taormina,Italy,Jul.2002 [46] Z.Fu,P.Zerfos,H.Luo,S.Lu,L.Zhang,M.Gerla,Theimpactofmultihopwire-lesschannelonTCPthroughputandloss,Proc.IEEEInfocom,SanFranciso,CA,USA,Mar.2003. [47] C.L.FullmerandJ.J.Garcia-Luna-Aceves,Solutionstohiddenterminalproblemsinwirelessnetworks,Proc.ACMSIGCOMM,Cannes,France,Sept.1997. [48] J.J.Garcia-Luna-AcevesandA.Tzamaloukas,Reversingthecollision-avoidancehandshakeinwirelessnetworks,Proc.ACM/IEEEMobicom,Seattle,Washington,USA,Aug.1999. [49] H.Garcia-Molina,Electionsinadistributedcomputingsystem,IEEETrans.Comp.,vol.31,no.1,Jan.1982. [50] M.Gerla,R.Bagrodia,L.Zhang,K.Tang,andL.Wang,TCPoverwirelessmul-tihopprotocols:simulationandexperiments,Proc.IEEEICC,Vancouver,BritishColumbia,Canada,Jun.1999. [51] M.Gerla,K.Tang,andR.Bagrodia,TCPperformanceinwirelessmultihopnet-works,Proc.IEEEWMCSA,Washington,DC,USA,Feb.1999. [52] S.Gobriel,R.Melhem,andD.Mosse,Auniedinterference/collisionanalysisforpower-awareadhocnetworks,inProc.Infocom,HongKong,China,March2004.

PAGE 338

[53] D.GrossandC.M.Harris,Fundamentalsofqueueingtheory,3rded.,JohnWiley&Sons,Inc,NewYork,1998 [54] X.Guo,S.Roy,andW.S.Conner,Spatialreuseinwirelessad-hocnetworks,inProc.VTC,Orlando,Florida,USA,2003. [55] P.GuptaandP.R.Kumar,Thecapacityofwirelessnetworks,IEEETransactionsonInformationTheory,vol.46,pp.388,2000. [56] R.Gupta,Z.Jia,T.Tung,andJ.Walrand,Interference-awareqosrouting(IQRout-ing)forad-hocnetworks,inIEEEGLOBECOM,St.Louis,Missouri,USA,No-vember2005. [57] V.Gupta,S.V.Krishnamurthy,andM.Faloutsos,ImprovingtheperformanceofTCPinthepresenceofinteractingUDPowsinadhocnetworks,IFIPNetworking,Athens,Greece,May2004. [58] Z.Hadzi-VelkovandB.Spasenovski,Saturationthroughput-delayanalysisofIEEE802.11DCFinfadingchannel,Proc.IEEEICC,Anchorage,Alaska,May,2003. [59] Z.J.HaasandJ.Deng,Dualbusytonemultipleaccess(DBTMA)-amultipleaccesscontrolforadhocnetworks,IEEETrans.Commun.,vol.50,pp.975-985,June2002. [60] Z.J.HaasandJ.Deng,Dualbusytonemultipleaccess(DBTMA):performanceresults,Proc.IEEEWCNC,NewOrleans,Louisiana,Sept.1999. [61] R.HekmatandP.V.Mieghem,Interferenceinwirelessmulti-hopad-hocnetworksanditseffectonnetworkcapacity,inMed-Hoc-Net,Sardegna,Italy,September2002. [62] T.Henderson,D.Kotz,andI.Abyzov,Thechangingusageofamaturecampus-widewirelessnetwork,inACMMobiCom,Philadelphia,PA,USA,Sept.2004. [63] M.Heusse,F.Rousseau,G.Berger-Sabbatel,andA.Duda,Performanceanomalyof802.11b,inIEEEInfocom,SanFranciso,CA,USA,March2003. [64] G.HollandandN.H.Vaidya,AnalysisofTCPperformanceovermobileadhocnetworks,Proc.ACMMobicom,Seattle,Washington,USA,Aug.1999. [65] G.Holland,N.Vaidya,andP.Bahl,Arate-adaptiveMACprotocolforwirelessnetworks,InACMMOBICOM,Rome,Italy,July2001. [66] T.S.HoandK.C.Chen,PerformanceanalysisofIEEE802.11CSMA/CAmediumaccesscontrolprotocol,Proc.IEEEPIMRC,Taipei,Taiwan,1996. [67] X.L.HuangandB.Bensaou,Onmax-minfairnessandschedulinginwirelessad-hocnetworks:analyticalframeworkandimplementation,ACMMobihoc,LongBeach,California,USA,Oct.2001.

PAGE 339

[68] [69] [70] [71] [72] [73] ITU-TG.114,One-waytransmissiontime,1996. [74] ITU-TG.1010,End-usermultimediaQoScategories,2001. [75] K.Jain,J.Padhye,V.Padmanabhan,andL.Qiu,Impactofinterferenceonmulti-hopwirelessnetworkperformance,inACMMobicom,SanDiego,CA,USA,Sep-tember2003. [76] Z.Ji,Y.Yang,J.Zhou,M.Takai,andR.Bagrodia,Exploitingmediumaccessdi-versityinrateadaptivewirelessLANs,inACMMobiCom,Philadelphia,PA,USA,Sept.2004. [77] Z.Jia,R.Gupta,J.Walrand,andP.Varaiya,Bandwidthguaranteedroutingforad-hocnetworkswithinterferenceconsideration,in10thIEEESymposiumonCom-putersandCommunications(ISCC),Cartagena,Spain,June2005. [78] V.Kanodia,C.Li,A.Sabharwal,B.Sadeghji,andE.Knightly,Distributedmulti-hopschedulingandmediumaccesswithdelayandthroughputconstraints,Proc.ACMMobiCom,Rome,Italy,2001. [79] S.Jamin,P.B.Danzig,S.Shenker,andL.Zhang,Ameasurement-basedadmissioncontrolalgorithmforintegratedservicepacketnetworks,IEEE/ACMTransactionsonNetworking,vol.5,no.1,Feb.1997. [80] A.KamermanandL.Monteban,WaveLANII:Ahigh-performancewirelessLANfortheunlicensedband,BellLabsTechnicalJournal,Summer1997. [81] V.Kanodia,C.Li,A.Sabharwal,B.Sadeghji,andE.Knightly,Distributedmulti-hopschedulingandmediumaccesswithdelayandthroughputconstraints,Proc.ofACMMobiCom2001,Rome,Italy,July2001. [82] P.Karn,MACA-anewchannelaccessmethodforpacketradio,inARRL/CRRLAmateurRadio9thComputerNetworkingConf,London,Ontario,Canada,1990,pp.134.

PAGE 340

[83] D.Katabi,M.Handley,andC.Rohrs,Congestioncontrolforhighbandwidth-delayproductnetworks,ACMSIGCOMM,Pittsburgh,PA,USA,2002. [84] V.KawadiaandP.R.Kumar,Acautionaryperspectiveoncross-layerdesign,IEEEWirelessCommunications,vol.12,no.1,pp.3,Feb.2005. [85] H.KimandJ.Hou,Improvingprotocolcapacitywithmodel-basedframeschedul-inginIEEE802.11-operatedWLANs,Proc.ACMMobiCom,SanDiego,CA,USA,Sep.2003. [86] L.Kleinrock,Queueingsystems,volumeI,JohnWiley&Sons,1975. [87] L.Kleinrock,Queueingsystems,volumeII,JohnWiley&Sons,1975. [88] A.KopselandA.Wolisz,Voicetransmissioninanieee802.11wlanbasedaccessnetwork,Proc.WoWMoM,Rome,Italy,2001. [89] H.T.Kung,T.BlackwellandA.Chapman,Credit-basedowcontrolforATMnetworks:creditupdateprotocol,adaptivecreditallocation,andstatisticalmulti-plexing,Proc.ACMSigcomm,London,UK,Sep.1994. [90] Y.Kwon,Y.Fang,andH.Latchman,AnovelMACprotocolwithfastcollisionres-olutionforwirelssLANs,Proc.IEEEINFOCOM,SanFranciso,CA,USA,2003. [91] J.Li,C.Blake,D.S.J.DeCouto,H.I.Lee,andR.Morris,Capacityofadhocwirelessnetworks,inProc.ACMMobiCom,Rome,Italy,July2001. [92] Y.Li,H.Wu,D.Perkins,N.Tzeng,andM.Bayoumi,MAC-SCC:mediumaccesscontrolwithaseparatecontrolchannelformultihopwirelessnetworks,23rdInter-nationalConferenceonDistributedComputingSystemsWorkshops(ICDCSW'03),Providence,RhodeIsland,USA,May2003. [93] Z.Li,S.Nandi,andA.K.Gupta,ImprovingMACperformanceinwirelessadhocnetworksusingenhancedcarriersensing(ECS),inThirdIFIPNetworking,Athens,Greece,2004. [94] A.Lindgren,A.Almquist,andO.Schelen,EvaluationofqualityofserviceschemesforIEEE802.11wirelessLANs,Proc.LocalComputerNetworks,Tampa,Florida,2001. [95] J.LiuandS.Singh,ATCP:TCPformobileadhocnetworks,IEEEJ.Sel.AreasCommun.vol.19,pp.1300-1315,Jul.2001. [96] S.C.Lo,G.Lee,andW.T.Chen,AnefcientmultipollingmechanismforIEEE802.11wirelessLANs,IEEETrans.Computers,vol.52,no.6,pp.764-778,2003. [97] S.Lu,T.Nandagopal,andV.Bharghavan,Awirelessfairservicealgorithmforpacketcellularnetworks,inACMMobiCom,Dallas,Texas,USA,Oct.1998.

PAGE 341

[98] H.LuoandS.Lu,andV.BharghavanAnewmodelforpacketschedulinginmul-tihopwirelessnetworks,Proc.ACMMobicom,Boston,Massachusetts,USA,Aug.2000. [99] S.Mangold,S.Choi,P.May,O.Klein,G.Hietz,andL.Stibor,IEEE802.11ewirelessLANforqualityofservice,Proc.EuropeanWireless'02,Florence,Italy,Feb.2002. [100] P.P.MishraandH.Kanakia,Ahopbyhoprate-basedcongestioncontrolscheme,Proc.ACMSigcomm,Baltimore,Maryland,USA,Aug.1992. [101] J.MoandJ.Walrand,Fairend-to-endwindow-basedcongestioncontrol,IEEE/ACMTransactionsonNetworking,vol.8,no.8,pp.556,Oct.2000. [102] J.Monks,V.Bharghavan,andW.Hwu,Apowercontrolledmultipleaccessprotocolforwirelesspacketnetworks,Proc.IEEEINFOCOM,Anchorage,Alaska,USA,April2001. [103] J.P.Monks,P.SinhaandV.Bharghavan,LimitationsofTCP-ELFNforadhocnetworks,Proc.MOMUC,Tokyo,Japan,Oct.2000. [104] A.MuqattashandM.Krunz,Powercontrolleddualchannel(PCDC)mediumac-cessprotocolforwirelessadhocnetworks,Proc.IEEEINFOCOM,SanFranciso,CA,USA,March2003. [105] T.Nandagopal,T.-E.Kim,X.Gao,andV.Bharghavan,AchievingMAClayerfair-nessinwirelesspacketnetworks,Proc.ACMMobicom,Boston,Massachusetts,USA,Aug.2000. [106] Thenetworksimulatorns-2.http://www.isi.edu/nsnam/ns. [107] W.Pattara-Atikom,P.Krishnamurthy,andS.Banerjee,DistributedmechanismsforqualityofserviceinwirelessLANs,IEEEWirelessCommunications,June2003. [108] C.Perkins,E.M.Royer,S.R.Das,andM.K.Marina,Performancecomparisonoftwoon-demandroutingprotocolsforadhocnetworks,IEEEPers.Commun.,vol.8,pp.16-28,Feb.2001. [109] H.PerrosandK.Elsayed,Calladmissioncontrolschemes:areview,IEEECom-municationsMagazine,Nov.1996. [110] S.Pilosof,R.Ramjee,D.Raz,Y.Shavitt,andP.Sinha,UnderstandingTCPfairnessoverWirelessLAN,Proc.IEEEINFOCOM,SanFranciso,CA,USA,March2003. [111] J.Postel,Internetcontrolmessageprotocol,IETFRFC792. [112] L.P.A.Robichaud,Signalowgraphsandapplications,Prentice-Hall,1962 [113] B.Sadeghi,V.Kanodia,A.Sabharwal,andE.Knightly,Opportunisticmediaaccessformultirateadhocnetworks,inACMMobiCom,SanDiego,CA,USA,Sept.2003.

PAGE 342

[114] S.-T.SheuandT.-F.Sheu,Abandwidthallocation/sharing/extensionprotocolformultimediaoverIEEE802.11adhocwirelessLANs,IEEEJSAC,vol.19,pp.2065-2080,Oct.2001. [115] M.ShreedharandG.Varghese,Efcientfairqueuingusingdecitround-robin,IEEE/ACKTrans.Netw.,vol.4,pp.375-385,Jun.1996. [116] S.SinghandJ.Kurose,Electing'good'leaders,J.Par.Distr.Comput.,vol.18,no.1,May1993. [117] S.SinghandC.S.Raghavendra,PAMAS-powerawaremulti-accessprotocolwithsignallingforadhocnetworks,ComputerCommunicationsReview,July1998. [118] J.SoandN.H.Vaidya,Amulti-channelMACprotocolforadhocwirelessnet-works,TechnicalReport,Jan.2003. [119] J.L.SobrinhoandA.S.Krishnakumar,Real-timetrafcovertheIEEE802.11mediumaccesscontrollayer,BellLabsTech.J.,pp.172-187,1996. [120] F.TalucciandM.Gerla,MACA-BI(MACAbyinvitation):areceiverorientedaccessprotocolforwirelessmultihopnetworks,Proc.IEEEPIMRC,Helsinki,Fin-land,Sep.1997. [121] G.TanandJ.Guttag,Time-basedfairnessimprovesperformanceinmulti-ratewire-lessLANs,inTheUSENIXAnnualTechnicalConference,Boston,Massachusetts,USA,June2004. [122] F.A.TobagiandL.Kleinrock,Packetswitchinginradiochannels:PartIIThehid-denterminalproblemincarriersensemultiple-accessandthebusy-tonesolution,IEEETrans.Commun.,vol.COM-23,pp.1417-1433,Dec.1975. [123] N.Vaidya,P.Bahl,andS.Gupta,DistributedfairschedulinginawirelessLAN,inACMMobiCom,Boston,Massachusetts,USA,Aug2000. [124] M.Veeraraghavan,N.Cocker,andT.Moors,Supportofvoiceservicesinieee802.11wirelesslans,InProc.INFOCOM,Anchorage,Alaska,USA,2001. [125] A.Veres,A.T.Campbell,M.Barry,andL.-H.Sun,Supportingservicedifferentia-tioninwirelesspacketnetworksusingdistributedcontrol,IEEEJSAC,vol.19,pp.2081-2093,Oct.2001. [126] M.A.VisserandM.E.Zarki,Voiceanddatatransmissionoveran802.11wirelessnetwork,Proc.IEEEPIMRC,Toronto,Ont.,Canada,1995. [127] F.WangandY.Zhang,ImprovingTCPperformanceovermobilead-hocnetworkswithout-of-orderdetectionandresponse,Proc.ACMMobihoc,Lausanne,Switzer-land,Jun.2002.

PAGE 343

[128] J.Wang,H.Zhai,W.Liu,andY.Fang,Reliableandefcientpacketforwardingbyutilizingpathdiversityinwirelessadhocnetworks,IEEEMilitaryCommunicationsConference(Milcom'04),Monterey,California,USA,Nov.2004. [129] J.Wang,H.Zhai,andY.Fang,Opportunisticmediaaccesscontrolandrateadap-tationforwirelessadhocnetworks,IEEECommunicationsConference(ICC'04),Paris,June,2004. [130] ,OpportunisticpacketschedulingandmediaaccesscontrolforwirelessLANsandmulti-hopadhocnetworks,IEEEWirelessCommunicationsandNetworkingConference(WCNC'04),Atlanta,March,2004. [131] J.Wang,H.Zhai,Y.Fang,J.M.Shea,andD.WuOMAR:UtilizingMultiuserDiversityinWirelessAdHocNetworks,toappearinIEEETransactionsonMobileComputing. [132] Y.WangandJ.J.Garcia-Luna-Aceves,Ahybridcollisionavoidanceschemeforadhocnetworks,WirelessNetworks,vol.10,pp.439-436,Jul.2004. [133] C.WuandV.Li,Receiver-initiatedbusy-tonemultipleaccessinpacketradionet-works,inSIGCOMM'87:ProceedingsoftheACMworkshoponFrontiersincom-putercommunicationstechnology,Stowe,Vermont,USA,Aug.1987. [134] H.Wu,Y.Peng,K.Long,S.Cheng,andJ.Ma,PerformanceofreliabletransportprotocoloverIEEE802.11wirelessLAN:analysisandenhancement,Proc.IEEEINFOCOM,NewYork,NY,USA,June,2002. [135] S.Wu,Y.TsengandJ.Sheu,Intelligentmediumaccessformobileadhocnetworkswithbusytonesandpowercontrol,IEEEJournalonSelectedAreasinCommuni-cations,Vol.18,pp.1647-1657,Sept.2000. [136] S.-L.Wu,C.-Y.Lin,Y.-C.Tseng,andJ.-P.Sheu,Anewmulti-channelMACpro-tocolwithon-demandchannelassignmentformobileadhocnetworks,Int'lSymp.onParallelArchitectures,AlgorithmsandNetworks(I-SPAN),Dallas,Texas,USA,2000. [137] Y.Xiao,H.Li,andS.Choi,ProtectionandguaranteeforvoiceandvideotrafcinIEEE802.11eWirelessLANs,Proc.IEEEINFOCOM,HongKong,China,2004. [138] K.Xu,M.Gerla,andS.Bae,HoweffectiveistheIEEE802.11RTS/CTShand-shakeinadhocnetworks?inProc.IEEEGlobeCom,Taipei,Taiwan,2002. [139] K.Xu,M.Gerla,L.Qi,andY.Shu,EnhancingTCPfairnessinadhocwirelessnetworksusingneighborhoodRED,Proc.ACMMobicom,SanDiego,CA,USA,Sep.2003. [140] S.XuandT.Safadawi,DoestheIEEE802.11MACprotocolworkwellinmultihopwirelessadhocnetworks?IEEECommun.Mag.,vol.39,pp.130-137,Jun.2001.

PAGE 344

[141] Q.XueandA.Ganz,ProportionalservicedifferentiationinwirelessLANswithspacing-basedchanneloccupancyregulation,inACMMultimedia,NewYork,NY,USA,Oct.2004. [142] X.YangandN.H.Vaidya,Onthephysicalcarriersenseinwirelessadhocnet-works,inProc.IEEEInfocom,Miami,FL,USA,March2005. [143] J.YeeandH.Pezeshki-Esfahani,UnderstandingwirelessLANperformancetrade-offs.CommsDesign.com,Nov.2002. [144] C.H.Yeh,Inbandbusytoneforrobustmediumaccesscontrolinpervasivenetwork-ing,inFourthAnnualIEEEInternationalConferenceonPervasiveComputingandCommunicationsWorkshops,Pisa,Italy,March2006. [145] J.Y.YehandC.Chen,SupportofmultimediaserviceswiththeIEEE802-11MACprotocol,Proc.IEEEICC,NewYork,NY,USA,2002 [146] Y.YiandS.Shakkottai,Hop-by-hopcongestioncontroloverawirelessmulti-hopnetwork,Proc.IEEEInfocom,HongKong,China,Mar.2004. [147] H.Zhai,X.Chen,andY.Fang,Alleviatingintra-owandinter-owcontentionsforreliableserviceinmobileadhocnetworks,inProc.IEEEMILCOM,Monterey,California,USA,Nov.2004. [148] ,AcalladmissionandratecontrolschemeformultimediasupportoverIEEE802.11wirelessLANs,TheFirstInternationalConferenceonQualityofServiceinHeterogeneousWired/WirelessNetworks(QShine'04),Dallas,Texas,USA,Oct.2004. [149] ,AcalladmissionandratecontrolschemeformultimediasupportoverIEEE802.11wirelessLANs,ACMWirelessNetworks,vol.12,no.4,pp.451-463,August2006 [150] ,HowwellcantheIEEE802.11wirelessLANsupportqualityofservice?IEEETransactiononWirelessCommunications,vol.4,no.6,pp.3084-3094,Nov.2005. [151] ,Rate-basedtransportcontrolformobileadhocnetworks,IEEEWirelessCommunicationsandNetworkingConference(WCNC'05),NewOrleans,March,2005. [152] ,WCCP:ImprovingtransportlayerperformanceinmultihopadhocnetworksbyexploitingMAClayerinformation,acceptedforpublicationinIEEETransac-tionsonWirelessCommunications. [153] H.Zhai,andY.Fang,Mediumaccesscontrolprotocolsinmobileadhocnetworks:problemsandsolutions,HandbookofTheoreticalandAlgorithmicAspectsofAdHoc,Sensor,andPeer-to-PeerNetworks,editedbyJ.Wu,pp.231-250,CRCPress,2005.

PAGE 345

[154] ,PerformanceofwirelessLANsbasedonIEEE802.11MACprotocols,Proc.ofIEEEInternationalSymposiumonPersonal,IndoorandMobileRadioCommu-nications(PIMRC'03),Beijing,China,Sep.2003. [155] ,Physicalcarriersensingandspatialreuseinmultirateandmultihopwire-lessadhocnetworks,inProc.ofTheIEEEInternationalConferenceonComputerCommunications(INFOCOM06),Barcelona,Spain,April23-29,2006. [156] ,Adistributedadaptivepacketconcatenationschemeforsensorandadhocnetworks,inProc.ofIEEEMilitaryCommunicationsConference(Milcom05),At-lanticCity,NewJersey,Oct.17-20,2005. [157] ,Distributedowcontrolandmediumaccesscontrolinmobileadhocnet-works,toappearinIEEETransactionsonMobileComputing. [158] ,Impactofroutingmetricsonpathcapacityinmulti-rateandmulti-hopwire-lessadhocnetworks,inProc.ofThe14thIEEEInternationalConferenceonNet-workProtocols(ICNP'06),SantaBarbara,California,November12-15,2006. [159] ,Asingle-channelsolutiontohidden/exposedterminalproblemsinwirelessadhocnetworks,submittedforpublication. [160] H.Zhai,Y.Kwon,andY.Fang,PerformanceanalysisofIEEE802.11MACpro-tocolsinwirelessLANs,WileyWirelessCommunicationsandMobileComputing,SpecialIssueonEmergingWLANTechnologiesandApplications,vol.4,pp.917-931,Dec.2004. [161] H.Zhai,J.Wang,X.Chen,andY.Fang,Mediumaccesscontrolinmobileadhocnetworks:Challengesandsolutions,invitedpaperinWileyWirelessCommunica-tionsandMobileComputing,SpecialIssueonAdHocWirelessNetworks,vol.6,issue2,pp.151-170,March2006. [162] H.Zhai,J.Wang,andY.Fang,Distributedpacketschedulingformultihopowsinadhocnetworks,inProc.IEEEWCNC,Atlanta,Georgia,USA,March,2004. [163] ,DUCHA:Adual-channelMACprotocolformobileadhocnetworks,toappearinIEEETransactionsonWirelessCommunications,2005. [164] ,ProvidingstatisticalQoSguaranteeforvoiceoveripintheIEEE802.11wirelessLANs,IEEEWirelessCommunicationMagazine(SpecialIssueonVoiceoverWirelessLocalAreaNetwork),vol.13,issue1,pp.36-43,Feb.2006.. [165] H.Zhai,J.Wang,Y.Fang,andD.Wu,Adual-channelMACprotocolformobileadhocnetworks,IEEEWorkshoponWirelessAdHocandSensorNetworks,inconjuctionwithIEEEGlobecom2004,Dallas,Texas,USA,Nov.2004. [166] H.Zhang,Servicedisciplinesforguaranteedperformanceserviceinpacket-switchingnetworks,ProceedingsoftheIEEE,vol.83(10),Oct.1995.

PAGE 346

[167] J.Zhu,X.Guo,L.L.Yang,andW.S.Conner,Leveragingspatialreusein802.11meshnetworkswithenhancedphysicalcarriersensing,inProc.IEEEICC,Paris,France,June2004.

PAGE 347

HongqiangZhaireceivedtheB.E.andM.E.degreesinelectricalengineeringfromTs-inghuaUniversity,Beijing,China,inJuly1999andJanuary2002respectively.HeworkedasaresearchinterninBellLabsResearchChinafromJune2001toDecember2001,inMicrosoftResearchAsiafromJanuary2002toJuly2002,andinKiyonInc.fromSeptem-ber2005toDecember2005.CurrentlyheispursuingthePhDdegreeintheDepartmentofElectricalandComputerEngineering,UniversityofFlorida.Hisresearchinterestsincludeperformanceanalysis,mediumaccesscontrol,qualityofservice,fairness,congestioncon-trol,routingalgorithmsandcross-layerdisigninwirelessnetworks.HeisastudentmemberofACMandIEEE. 328


Permanent Link: http://ufdc.ufl.edu/UFE0015668/00001

Material Information

Title: Cross-Layer Design of Networking Protocols in Wireless Local Area Networks and Mobile Ad Hoc Networks
Physical Description: Mixed Material
Copyright Date: 2008

Record Information

Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
System ID: UFE0015668:00001

Permanent Link: http://ufdc.ufl.edu/UFE0015668/00001

Material Information

Title: Cross-Layer Design of Networking Protocols in Wireless Local Area Networks and Mobile Ad Hoc Networks
Physical Description: Mixed Material
Copyright Date: 2008

Record Information

Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
System ID: UFE0015668:00001


This item has the following downloads:


Full Text











CROSS-LAYER DESIGN OF NETWORKING PROTOCOLS IN WIRELESS LOCAL
AREA NETWORKS AND MOBILE AD HOC NETWORKS















By

HONGQIANG ZHAI


A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY

UNIVERSITY OF FLORIDA


2006

































Copyright 2006

by

Hongqiang Zhai














Dedicated to my beloved parents and brothers.















ACKNOWLEDGMENTS

First and foremost, I would like to express my sincere gratitude to my advisor, Pro-

fessor Yuguang Fang, for his invaluable advice, encouragement and motivation during the

course of this work. This dissertation would not have been possible without his guid-

ance and support. I also thank him for his philosophical advice on both my academic and

nonacademic life, which made me more mature, scholastically and personally.

I thank Professors Shigang Chen, Jose Fortes, Pramod Khargonekar and Sartaj Sahni

for serving on my supervisory committee and for their valuable suggestions and construc-

tive criticism. Thanks also go to Prof. John Shea, Prof Tan Wong and Prof. Dapeng Wu,

for their many constructive suggestions and advice.

Many thanks are due to my colleagues Dr. Xiang Chen and Jianfeng Wang for their

collaboration. I also thank Dr. Younggoo Kwon, Dr. Wenjing Lou, Dr. Wenchao Ma, Dr.

Wei Liu, Dr. Byung-Seo Kim, Dr. Xuejun Tian, Dr. Sungwon Kim, Dr. Jae Sung Lim, Yu

Zheng, Yanchao Zhang, ShushanWen, Xiaoxia Huang, Yun Zhou, Jing Zhao, Chi Zhang,

Frank Goergen, Pan Li, Feng Chen, Shan Zhang, Rongsheng Huang and many others at

University of Florida for the years of friendship and many helpful discussions.

Last but not least, I owe a special debt of gratitude to my parents and my brothers.

Without their selfless love and support, I would never imagine what I have achieved.















TABLE OF CONTENTS
page

ACKNOW LEDGMENTS ...................... . ........ iv

LIST OF TA BLES . . . . . . . . xii

LIST OF FIGURE S . . . . . . . . xiii

A B STR A CT . . .. . . . . . . .. xviii

CHAPTER

1 INTRODUCTION .............................. 1

1.1 M otivation . . . . . . . . 1
1.2 Organization of the Dissertation ............ . .. 2

2 PERFORMANCE OF THE IEEE 802.11 DCF PROTOCOL IN WIRELESS
LAN S ........................... ...... 8

2.1 Introduction . . . . . . . . 8
2.2 Prelim inaries . . . . . . . 10
2.2.1 Distributed Coordination Function (DCF) ..... 10
2.2.2 System M odeling ........................ 11
2.3 The Probability Distribution of the MAC Layer Service Time . 12
2.3.1 MAC Layer Service Time ... . . . .... 12
2.3.2 Probability Generating Functions (PGF) of MAC Layer Service
Tim e . . ... .... .... . ... 13
2.3.3 The Processes of Collision and Successful Transmission . 15
2.3.4 Decrement Process of Backoff Timer . . . 16
2.3.5 Markov Chain Model for the Exponential Backoff Procedure 17
2.3.6 Generalized State Transition Diagram . . . 18
2.3.7 Probability Distribution Modeling . . . . 20
2.3.8 Derivation of Transmission Probability . . 23
2.4 Queueing Modeling and Analysis . . . . . 25
2.4.1 Problem formulation. . . . 25
2.4.2 The steady-state probability of the M/G/1/K queue . . 26
2.4.3 Conditional Collision Probability pc and Distribution of MAC
Layer Service Time . . . . . 27
2.4.4 Performance Metrics of the Queueing System . . 27
2.4.5 Throughput . . . . . . . 27
2.4.6 Numerical Results. .................. .. 28










2.5 Performance Evaluation ............
2.5.1 Simulation Environments .. ..................
2.5.2 Probability Distribution of MAC Layer Service Time ......
2.5.3 Comparison of M/G/1/K and M/M/1/K Approximations with
Simulation Results ...........
2.6 C conclusions . . . . . . . .

3 HOW WELL CAN THE IEEE 802.11 DCF PROTOCOL SUPPORT QOS IN
WIRELESS LANS ................


3.1 Introduction . .......
3.2 Preliminaries . . .
3.2.1 Operations of the IEEE 802.11 . ......
3.2.2 Related W ork . ............
3.3 Analytical Study of the IEEE 802.11 . . .
3.3.1 Maximum Throughput and Available Bandwidth .
3.3.2 Delay and Delay Variation . ..
3.3.3 Packet Loss Rate . . . . .
3.4 Simulation Study of the IEEE 802.11 . ......
3.4.1 Simulation Configuration . . . .
3.4.2 Simulation Results . . . .
3.5 Discussions ......... . . .
3.5.1 Impact of Fading Channel . . . .
3.5.2 Impact of Prioritized MAC . ........


3.6 C conclusion . . . . . .


. . 35
. . 37
. . 37
. . 38
. . 40
. . 40
. . 47
. . 54
. . 56
. . 56
. . 58
. . 60
. . 60
. . 6 1
. . 6 1


4 A CALL ADMISSION AND RATE CONTROL SCHEME FOR MULTIME-
DIA SUPPORT OVER IEEE 802.11 WIRELESS LANS .....

4.1 Introduction .. ... ... ... ... .
4.2 Background ...............
4.2.1 Operations of the IEEE 802.11 DCF Protocol ..........
4.2.2 QoS Requirements for Multimedia Services ...........
4.3 Channel Busyness Ratio ..........................
4.3.1 Definition of Channel Busyness Ratio .............
4.3.2 Channel busyness ratio: an accurate sign of the network utilization
4.3.3 Measurement of Channel Busyness Ratio . . . .
4.4 CARC: Call Admission and Rate Control .. .............
4.4.1 D esign Rationale . . . . . . .
4.4.2 Call Admission Control .. ..................
4.4.3 R ate C control . . . . . . .
4.5 Performance Evaluation of CARC .........
4.5.1 Simulation Configuration ..........
4.5.2 Simulation Results . . . . . .
4.6 D discussions . . . . . . . .
4.6.1 Impact of Fading Channel . . . . . .










4.6.2 Impact of Prioritized MAC . .
4.7 C conclusion . . . . . . . .

5 DISTRIBUTED FAIR AND EFFICIENT RESOURCE ALLOCATION WITH
QOS SUPPORT OVER IEEE 802.11 WLANS .......


5.1 Introduction . . . . . .
5.2 Design Rationale . ................
5.2.1 Efficiency and QoS . ...
5.2.2 Fairness . .
5.3 Distributed Resource Allocation (DRA) .
5.3.1 Basic Framework . ............
5.3.2 Fairness Support . . . ..
5.3.3 QoS Support . . . . .
5.3.4 Multiple Channel Rates Support .
5.4 Convergence Analysis. . . . .....
5.4.1 Convergence of Multiplicative-Increase Phase .
5.4.2 Convergence to Fairness Equilibrium . ..
5.4.3 D discussion . . . . .
5.4.4 Parameter Selection . ..........
5.5 Performance Evaluation . . . ..
5.5.1 Simulation Setup . . ........
5.5.2 Channel Busyness Ratio Threshold . ..
5.5.3 Fairness ............... . .
5.5.4 Efficiency, Delay and Collision . ..
5.5.5 Quality of Service . . . ..
5.6 Related Work and Discussions . .........
5.7 Conclusions . . . . . .

6 PHYSICAL CARRIER SENSING AND SPATIAL REUSE IN
AND MULTIHOP WIRELESS AD HOC NETWORKS .

6.1 Introduction . . . . . .
6.2 Optimum Carrier Sensing Range . . ..


. 88
. 92
. . 92
. . 94
. . 95
. . 96
. . 100
. . 100
. . 102
. . 102
. . 102
. . 105
. . 109
. . 110
. . 110
. . 110
. . 1 1 1
.... .. 112
. . 115
. . 116
. . 119
. . 12 1

MULTIRATE


6.2.1 Aggregate Throughput and SINR at the Worst Case . .
6.2.2 Maximum Throughput and Optimum Carrier Sensing Range un-
der Shannon Capacity . . .
6.2.3 Maximum Throughput and Optimum Carrier Sensing Range un-
der the Discrete Channel Rates of the IEEE 802.11 . .
6.2.4 Impact of Random Topology . . .
6.2.5 Tradeoff between Exposed Terminal Problem and the Hidden
Terminals Problem . . . .
6.2.6 Carrier Sensing Range and Strategies for Bidirectional Hand-
shakes ...... . . ......
6.2.7 Optimum Carrier Sensing Range . . .....


. 86
. 87


. 127

. 130

. 131
. 133

. 134

S136
. 140









6.3 Utilize Multirate Capability of 802.11 in Wireless Multihop Ad Hoc
N etw orks ... .. . . . . . ... 140
6.3.1 How to Set the Carrier Sensing Threshold for Multirate 802.11
MAC protocol . . . . . . 140
6.3.2 How to Choose Next Hops, Channel Rates and Set the Carrier
Sensing Threshold for Multihop Flows . . ..... 142
6.4 Simulation Studies . . . . . . . 149
6.4.1 NS2 Extensions and Simulation Setup . 150
6.4.2 Optimum Carrier Sensing Range . . ..... 150
6.4.3 Spatial Reuse and End-to-End Performance of Multihop Flows 153
6.5 Conclusions . . . . . . . . 154

7 A DUAL-CHANNEL MAC PROTOCOL FOR MOBILE AD HOC NETWORKS 156

7.1 Introduction . . . . . . . . 156
7.2 Background .. . . . . . . . 160
7.2.1 Physical M odel . . . . . . 160
7.2.2 Transmission Range and Sensing/Interference Range . 160
7.3 Problems and The Desired Protocol Behavior . 161
7.3.1 Hidden and Exposed Terminal Problem ... . ...... 161
7.3.2 Limitations of NAV Setup Procedure . . 162
7.3.3 Receiver Blocking Problem . . . . . 163
7.3.4 Intra-Flow Contention . . . . . 164
7.3.5 Inter-flow Contention . . . . . 165
7.3.6 The Desired Protocol Behavior ... . . ...... 165
7.3.7 Limitation of IEEE 802.11 MAC Using Single Channel . 166
7.4 DUCHA: A New Dual-Channel MAC Protocol . 166
7.4.1 Protocol Overview . . . . . . 166
7.4.2 Basic Message Exchange . . . . . 167
7.4.3 Solutions to the Aforementioned Problems . . . 169
7.4.4 Remarks on the proposed protocol... . . . 171
7.5 Performance Evaluation . . . . . . 172
7.5.1 Simulation Environments . . . . . 172
7.5.2 Simple Scenarios . . . . . . 173
7.5.3 Random Topology for One-hop Flows . 176
7.5.4 Random Topology for Multihop Flows ... . ...... 178
7.6 Conclusions . . . . . . . . 181

8 A SINGLE-CHANNEL SOLUTION TO HIDDEN/EXPOSED TERMINAL
PROBLEMS IN WIRELESS AD HOC NETWORKS .. . 183

8.1 Introduction . . . . . . . 183
8.2 Various Ranges in Wireless Multihop Ad Hoc Networks . . 188
8.3 Addressing the Hidden/Exposed Terminal Problems with Short Busy
Advertisement Signal .. . . . . 189
8.3.1 Basic Operations in the SBA Procedure ... . ...... 190









8.3.2 Mitigating Exposed Terminal Problem by Adjusting Carrier Sens-
ing R ange . . . . . . . 191
8.3.3 Parameters in SBA Procedure . . . . 191
8.3.4 Positions of IDFS Periods in the DATA Frame . . 193
8.3.5 Busy Advertisement Signal . . . . . 195
8.3.6 Power Control for Short Busy Advertisement . . 195
8.3.7 Start and Stop SBA Procedure . . . . 196
8.3.8 Synchronization Issue . . . . . 198
8.3.9 Accumulative Acknowledgement . . 198
8.3.10 CTS Dominance ...... . . . . 199
8.3.11 Compatibility with Legacy 802.11 MAC Scheme . . 199
8.4 Maximize Spatial Reuse Ratio and Minimize Power Consumption by
Power Control ....... . . . . ..... 199
8.4.1 Power Control for Both DATA Frame and Busy Advertisement
in SBA-M AC .. . . . . 200
8.4.2 Power Control for the Approach Using A Large Carrier Sensing
Range ... ...... ............... 202
8.5 Performance Analysis . . . . . . .204
8.5.1 Spatial Reuse Ratio . . . . . .204
8.5.2 Protocol Overhead. . . . . . .204
8.5.3 Numerical Results . . . . . . .206
8.6 Conclusions . . . . . . . .208

9 A DISTRIBUTED PACKET CONCATENATION SCHEME FOR SENSOR
AND AD HOC NETWORKS . . . . . . 211

9.1 Introduction . . . . . . . .211
9.2 Operations of the IEEE 802.11 . . . 213
9.3 Adaptive Packet Concatenation (APC) Scheme and Performance Analysis214
9.3.1 Basic Schem e . .. . . . . 214
9.3.2 Performance Analysis of the Network Throughput in the Single
H op Case ... .. . .. . . . .217
9.3.3 Performance Analysis of the Network Throughput in a Multihop
Network . . . . . . .221
9.4 Conclusion . . . . . . . .225

10 IMPACT OF ROUTING METRICS ON PATH CAPACITY IN MULTIRATE
AND MULTIHOP WIRELESS AD HOC NETWORKS . . 226

10.1 Introduction . . . . . . . .226
10.2 Impact of Multirate Capability on Path Selection In Wireless Ad Hoc
N etw orks . ... .. . .. .. . .231
10.2.1 Receiver Sensitivity and SNR for Multiple Rates . . 231
10.2.2 Tradeoff between the rate and the transmission distance . 232
10.2.3 Carrier Sensing Range, Interference and Spatial Reuse . 232
10.2.4 Effective Data Rate and Protocol Overhead . . . 233










10.3 Path Capacity in Wireless Ad Hoc Networks . . 234
10.3.1 Link Conflict Graph .... . . . . .235
10.3.2 Upper Bound of Path Capacity in Single Interference Model 237
10.3.3 Exact Path Capacity in Single Interference Model . . 240
10.3.4 Path Capacity in Multi-Interference Model with Variable Link
Rate ....... .. ... ......... .. .... 242
10.3.5 Extended to Multiple Paths between a Source and Its Destina-
tion or between Multiple Pairs of Source and Destination 243
10.3.6 Consider the packet error rate over each link in the link schedul-
ing algorithm . . . . . . .244
10.4 Path Selection in Wireless Ad Hoc Networks . . 244
10.4.1 Optimal Path Selection . . . . . .245
10.4.2 Using Routing Metrics in Path Selection . . 246
10.5 Performance Evaluation . . . . . . .247
10.5.1 Simulation Setup . . . . . . .247
10.5.2 Compared with Optimal Routing . . 248
10.5.3 Performance Evaluation of Six Routing Metrics in a Larger Topol-
ogy . ... . . . .. . .249
10.5.4 Path Capacity of a Single-Rate Network . . 252
10.6 Conclusions . . . . . . . .253


11 DISTRIBUTED FLOW CONTROL AND MEDIUM ACCESS CONTROL IN
MOBILE AD HOC NETWORKS . . .


Introduction ..... . . .. .. .. ..
Impact of MAC Layer Contentions on Traffic Flows . . .
OPET: Optimum Packet Scheduling for Each Traffic Flow . .
11.3.1 Overview . . . .....
11.3.2 Rule 1: Assigning High Channel Access Priority to Receivers


11.3.3 Rule 2: Backward-Pressure Scheduling .
11.3.4 Rule 3: Source Self-Constraint Scheme .
11.3.5 Rule 4: Round Robin Scheduling . .
11.4 Performance Evaluation . .........
11.4.1 Simple Scenarios . .........
11.4.2 Random Topology . ......
11.4.3 Random Topology with Mobility . .
11.4.4 Simulation results for TCP traffic . .
11.4.5 Notes on the relative benefits of the four tec
11.5 Related Works and Discussion . ......
11.6 Conclusions . . . . .


. 255


.255
. 258
. 261
. 261
. 261


. ... . .263
. .. . .268
. . . .270
. . . 2 7 1
. . . .272
. . . 2 73
. . . .276
. . . .277
hniques . 279
. . . 2 80
. . . .2 82


12 WCCP: IMPROVING TRANSPORT LAYER PERFORMANCE IN MULTI-
HOP AD HOC NETWORKS BY EXPLOITING MAC LAYER INFOR-
M A T IO N . . . . . . .. .

12.1 Introduction ...................... . .....


11.1
11.2
11.3


283

283










12.2 Medium Contention and Its Impact . . 286
12.2.1 TCP Performance Degradation Due to Coupling of Congestion
and Medium Contention . . . . .286
12.2.2 Optimal Congestion Window Size for TCP and Ideal Sending Rate288
12.2.3 Unfairness Problem Due to Medium Contention . . 290
12.3 Wireless Congestion Control Protocol (WCCP) . . . 292
12.3.1 Channel Busyness Ratio: Sign of Congestion and Available Band-


w idth . . .. . .
12.3.2 Measurement of Channel Busyness Ratio in Multihop Ad Hoc


292


Networks . .....
12.3.3 Inter-node Resource Allocation .
12.3.4 Intra-node Resource Allocation .
12.3.5 End-to-End Rate-Based Congestion
12.4 Performance Evaluation . .....
12.4.1 Chain Topology . ......
12.4.2 Random Topology . ....
12.5 Conclusions . ... ... ... ..


Control Scheme .


. .294
. . 295
. . 297
. . 299
. . 302
. . 303
. . 308
. . 308


13 CONCLUSIONS AND FUTURE WORK . ..........

13.1 Fairness in Mobile Ad Hoc Networks . ........
13.2 Quality of Service in Mobile Ad Hoc Networks . ...

REFERENCES . ...........

BIOGRAPHICAL SKETCH . ....................


310


. . 310
. . 313

. . 315


328















LIST OF TABLES
Table page

2-1 IEEE 802.11 system parameters . . . . . . 22

2-2 Saturation value of collision probability ................ 22

3-1 QoS requirements for multimedia services . . . . 36

3-2 IEEE 802.11 system parameters . . . . . . 42

4-1 IEEE 802.11 system parameters . . . . . . 71

4-2 The mean, standard deviation (SD), and 97'th, 99'th, 99.9'th percentile de-
lays (in seconds) for voice and video in the infrastructure mode. .. . 83

4-3 The mean, standard deviation (SD), and 97'th, 99'th, 99.9'th percentile de-
lays (in seconds) for voice and video in the ad hoc mode. . . 85

6-1 Signal-to-noise ratio and receiver sensitivity . . . . 131

7-1 Default values in the simulations . . . . . . 172

9-1 IEEE 802.11 system parameters . . . . . .220

10-1 Signal-to-noise ratio and receiver sensitivity . . . . .232

10-2 Run time of different routing algorithms . . . . . 253

12-1 Simulation results for TCP and UDP flows . . . . 289

12-2 Performance of WCCP and TCP in chain topology of Fig. 12-3(a) . 303















LIST OF FIGURES
Figure page

2-1 RTS/CTS mechanism and basic access mechanism of IEEE 802.11 . 11

2-2 Generalized state transition diagram of one example . . . 15

2-3 Generalized state transition diagram for transmission process . 19

2-4 Probability distribution of MAC layer service time ... . ...... 21

2-5 PDF of service time . . . . . . . 23

2-6 M ean of service time . . . . . . . 23

2-7 Queue characteristics . . . . . . . 28

2-8 MAC layer packet service time . . . . . 30

2-9 Comparisons between M/G/1/K, M/M/1/K models and simulation . 31

2-10 Average waiting time in non-saturated status ....... . 32

2-11 Average MAC layer service time . . . . . 33

3-1 Channel busyness ratio and utilization . . . . . 41

3-2 Collision probability and maximum normalized throughput with RTS/CTS
and payload size of 8000bits . . . . . 45

3-3 Impact of payload size and the RTS/CTS mechanism . ..... 47

3-4 Mean and standard deviation of service time . . . . 49

3-5 Packet delay . . . . . . . . 54

3-6 Simulation results when payload size = 8000bits . . . 57

3-7 Simulation results when n=50 and payload size = 8000bits . . 58

3-8 Simulation results when n=50 and payload size = 8000bits . . 60

4-1 Channel busyness ratio and utilization . . . . . 70

4-2 Simulation results when number of nodes equals 50 and RTS/CTS mech-
anism is used . . . . . . . 71









4-3 Infrastructure mode: the number of real-time and TCP flows increases
over time. Channel rate is 2 Mbps . . . . . 82

4-4 End-to-end delay of all voice and video packets in infrastructure mode 83

4-5 Ad hoc mode: the number of real-time and TCP flows increases over time.
Channel rate is 2 M bps . . . . . . 84

4-6 End-to-end delay of all voice and video packets in ad hoc mode . 85

4-7 Channel utilization in ad hoc mode . . . . . 86

5-1 Maximum and saturated throughput with different number of nodes (RTS/CTS
is used, packet length = 1000bytes, channel rate = 11Mbps) . 94

5-2 Convergence speed of multiplicative-increase phase (packet length = 1000bytes,
channel rate = 11M bps) . . . . . . 105

5-3 Convergence speed of AIMD phases when 6 = 0.5 . ..... 109

5-4 Impact of payload size L and number of nodes n on the optimal threshold
for channel busyness ratio brh . . . . . 111

5-5 Fairness convergence with RTS/CTS: one greedy node joins the network
every 10 seconds (packet length = 1000bytes, each point is averaged
over 1 second) . . . . . . . 113

5-6 Max-min fairness under different traffic rates (packet length = 1000bytes) 114

5-7 DRA: fairness with multiple channel bit rates (RTS/CTS is used) . 115

5-8 802.11: fairness with multiple channel bit rates (RTS/CTS is used) . 115

5-9 Throughput, MAC delay and collision probability with RTS/CTS . 117

5-10 QoS performance inDRA ................... .118

5-11 QoS performance in 802.11 . . . . . . 118

6-1 Interference m odel . . . . . . . 128

6-2 Carrier sensing threshold with Shanon Capacity . 130

6-3 Carrier sensing threshold with different SINR . . . 131

6-4 Carrier sensing threshold with discrete channel rates of 802.11 . 132

6-5 Tradeoff between exposed terminal problem and hidden terminal problem 134

6-6 Large carrier sensing range with carrier sensing strategy II for CTS/ACK 139

6-7 Multiple carrier sensing thresholds may result in collisions . . 141









6-8 Bandwidth distance product . . . . . . 144

6-9 Maximum end-to-end throughput for different hop distance . . 145

6-10 Spatial reuse ratio for multihop flows (a) at worst case, (b) in a single chain
topology with one way traffic . . . . . 147

6-11 Optimum carrier sensing threshold for one-hop flows . . . 152

6-12 Optimum carrier sensing threshold for multi-hop flows . . .. 152

7-1 A simple scenario to illustrate the problems . . 162

7-2 Chain topology . . . . . . . . 165

7-3 Proposed protocol . . . . . . . 167

7-4 One simple topology . . . . . . . 173

7-5 Simulation results for the simple topology . . . 174

7-6 End-to-end throughput for the 9-node chain topology . .. . 177

7-7 Simulation results for random one-hop flows with different minimum one
hop distance . . . . . . . . 177

7-8 Simulation results for multihop flows in random topology . . 179

8-1 Hidden terminal problem . . . . . . 184

8-2 Carrier sensing range and interference range in LCS and SBA-MAC . 185

8-3 Four-way handshake with busy advertisement signals . .. . 190

8-4 Positions of IDFS periods in the DATA frame . . 193

8-5 Power control in SBA-MAC . . . . . . 200

8-6 Occupied area for a transmission normalized over the communication ra-
dius (PC: power control for DATA frames)... . . .207

8-7 Occupied area for a transmission normalized over the communication ra-
dius w hen dh = d . . . . . . . 207

8-8 Channel time for a transmitted packet . . . . . 208

8-9 Channel time for a transmitted packet . . . . . 209

8-10 Performance gain of SBA-MAC compared to the approach using a large
carrier sensing range and the FAMA scheme . . 209

9-1 RTS/CTS mechanism and basic access mechanism of IEEE 802.11 . 214










Protocol stack ..................... . .....

The super packet structure.... . . .....

Throughput when channel rate is 1Mbps, Lth = 2346bytes and RTS/CTS


. . ... . . 2 2 0


9-5 Throughput when channel rate is 1, 2, 5.5 and 11Mbps
mechanism is used ...............

9-6 Chain topology ..................

9-7 Maximum end-to-end throughput of a multihop flow ..

9-8 Maximum end-to-end throughput of a multihop flow ..

10-1 Paths between the source S and the destination D .

10-2 A five-link chain topology and its link Conflict graph .

10-3 A path with an odd cycle in the link conflict graph .

10-4 Path capacity for different routing algorithms . ..

10-5 Path capacity for different routing algorithms . ..

10-6 Path lengths for different routing algorithms . ...

10-7 Source-destination distance . ............

10-8 Path capacity solving time . ............

10-9 Path capacity for a single rate network . .

11-1 Chain topology and cross topology . ........

11-2 TCP performance in a 9-node chain topology . ..

11-3 Optimum packet scheduling for chain topology . .

11-4 The packet format of RTSM and CTSR . .....

11-5 The algorithms of backward-pressure scheme . .

11-6 Message sequence for packet transmission . ....

11-7 The packet scheduling for resolving congestion . .

11-8 Simulation results for the 9-node chain topology (Fig. 1
topology (Fig. 11-1(b)) . ............

11-9 Simulation results for the random topology . ...


and RTS/CTS
S. . 221

S. . 222

S. . 224

S. . 224

S. . 230

S. . 235

S. . 239

S. . 249

S. . 250

S. . 251

S. . 251

S. . 252

S. . 254

S. . 259

S. . 260

S. . 263

S. . 266

S. . 267

S. . 268

S. . 269

1-3) and cross
S. . 272

S. . 274


mechanism is used.


. 215

. 215









11-10 Simulation results for the random topology with mobility . . .277

11-11 Simulation results for the TCP traffic . . . . . 278

11-12 Grid topology with 16 TCP flows . . . . . 279

12-1 Chain topology with 9 nodes. Small circles denote the transmission range,
and the large circles denote the sensing range . . 286

12-2 Simulation results for 9-node chain topology ............... 287

12-3 Nine-node chain topology with different traffic distribution . ... 291

12-4 The relationship between channel busyness ratio and other metrics . 293

12-5 Rate control mechanism ................... . ...... 300

12-6 Simulation results for the nine-node chain topology with one flow . .304

12-7 Performance of scenario Fig.12-3(b) .................. 305

12-8 Performance of scenario Fig.12-3(c) .................. 306

12-9 Simulation results for random topology with precomputed paths: (a) min-
imum flow throughput in 20 runs, (b) minimum flow throughput aver-
aged over 20 runs, (c) maximum flow throughput averaged over 20 runs,
(d) ratio of averaged maximum flow throughput to averaged minimum
flow throughput. . . . . . . 307

12-10 Simulation results averaged over 20 runs in the random topology: (1) ag-
gregate throughput (Mbps), (2) fairness index, (3) end-to-end delay (s). 308

13-1 An original topology and its flow contention graph . . ..... 311


xvii















Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy

CROSS-LAYER DESIGN OF NETWORKING PROTOCOLS IN WIRELESS LOCAL
AREA NETWORKS AND MOBILE AD HOC NETWORKS

By

Hongqiang Zhai

August 2006

Chair: Yuguang "Michael" Fang
Major Department: Electrical and Computer Engineering

This Ph.D. dissertation focuses on design and analysis of efficient networking proto-

cols in wireless local area networks and ad hoc networks. Known as WI-FI technology,

wireless local area networks have become very popular today as an easy way of wire-

less access to the Internet. Wireless ad hoc networks also find a lot of applications which

need wireless access or require a low cost or an immediate deployment of networked sys-

tems, like battlefield communications, public safety networks, disaster rescues, and wire-

less metropolitan area networks. However, it is a very challenging task to design efficient

networking protocols to provide quality of service (QoS) and reliability in these networks.

Compared to wired networks, in wireless networks links are not independent any more;

bandwidth, power and processing ability are limited; channel errors happen frequently;

network topology is subject to constant change; and the network is often self-organized

and distributed. These challenges lead to close coupling among various layers in the proto-

col stack and a complete different medium access control (MAC) layer, and hence call for

a cross-layer design between the MAC layer and other layers.

The dissertation conducts a thorough theoretical study on a contention-based MAC

standard, IEEE 802.11, and investigates the close coupling between various protocol layers


xviii









and the MAC layer, features of which have been used to provide QoS and reliability, and

design efficient MAC, routing and transport protocols. The theoretical results discover that

the contention-based IEEE 802.11 MAC standard can well support Quality of Service and

at the same time achieve maximum aggregate throughput by regulating the access traffic.

Guided by the theoretical studies, the protocol design demonstrates various novel ways of

cross-layer design and their great benefit in improving performance of wireless networks.

Unlike prior research on cross-layer design approach in wireless networks which focused

on pure theoretical studies and either are too complicated to solve or the resulting solutions

become too simple to be practical because of many unpractical assumptions, the theoretical

studies and protocol design in this dissertation are based on the widely used IEEE 802.11

standard and hence can achieve immediate impact on products and revolutionize the way

that people design networked systems.















CHAPTER 1
INTRODUCTION

1.1 Motivation

With a rapid development in wireless communication technologies and the prolifer-

ation of mobile communication and computing devices like cell phones, PDAs and lap-

tops, wireless local area Networks (WLANs) and mobile ad hoc networks (MANETs)

have emerged as important parts of the envisioned future ubiquitous communication. In

recent years, the IEEE 802.11 wireless LAN has been increasingly employed to access the

Internet because of its simple deployment and low cost. MANETs are finding a variety

of applications such as disaster rescue, battlefield communications, inimical environment

monitoring, and collaborative computing. The widely studied sensor networks are special

applications of MANETs.

However, there are a lot of challenges for the networking protocols to work efficiently

in WLANs and MANETs. Unlike wired networks, some unique characteristics of WLANs

and MANETs seriously deteriorate performance of the networking protocols. These char-

acteristics include the time-varying channels due to path loss, fading and interference, the

vulnerable shared media access due to random access collision and the limited battery en-

ergy. In MANETs, the network topology may experience continuous change and cause fre-

quent route breakages and re-routing activity. And MANETs by nature are self-organized,

self-controlled and distributed. In other words, there is no centralized controller that has

perfect knowledge of all the nodes in the network. Instead, each node can only have incom-

plete or sometimes skewed view of the network. As a result, it has to make decisions with

imperfect information. Due to all these hurdles posed by WLANs and MANETs, simple,

efficient, fair, and energy-efficient networking protocols, while highly desirable, are not an

easy task.









These challenges call for the cross-layer design of the networking protocols in WLANs

and MANETs. For example, by scheduling the node with good channel quality to access

the channel, medium access control (MAC) protocols can achieve higher throughput. The

traditional congestion control protocol for the Internet, TCP, takes any packet loss as a

congestion sign. However, packet loss can be attributed to poor channel quality or route

failure due to mobility. It can achieve better performance if the TCP source can differ-

entiate the different reasons of packet losses by obtaining information from the routing

protocols and the physical and MAC layers. Routing protocols can also avoid unnecessary

re-routing messages if they can distinguish the packet losses for medium collision instead

of mobility. As to quality of service (QoS) and fairness, they seem to be formidable tasks

considering the unreliable physical channel, medium collisions and dynamically changing

network topology and traffic load. Cross-layer design seems a must to provide node-based

and flow-based fairness and end-to-end QoS guarantee.

1.2 Organization of the Dissertation

In this dissertation, we first conduct performance analysis of the Distributed Coordi-

nation Function (DCF) protocol in the IEEE 802.11 MAC standard in Chapter 2. We pro-

pose a new model using the signal transfer function of generalized state transfer diagram to

characterize the probability generation function of the medium access delay. With the prob-

ability distribution of medium access delay and queueing theory, most of the performance

metrics, such as throughput, delay, packet loss rate and various queue characteristics, can

be analyzed for the WLANs. Our results show that at the non-saturated state (i.e., each

node does not contend for the channel all the time and the total traffic rate does not exceed

the network capacity), the performance is dependent on the total traffic and almost indiffer-

ent to the number of transmitting stations. At the saturated state (i.e., each node has enough

traffic to keep contending the shared wireless channel), the number of transmitting stations

affects the performance more significantly.









In Chapter 3, we further derive the maximum throughput of the IEEE 802.11 DCF

protocol and accurate estimates for delay and delay variations in wireless LANs based on

our work in Chapter 2. We show that, by controlling the total traffic rate, the original 802.11

DCF protocol can support strict QoS requirements, such as those required by voice over IP

or streaming video, and at the same time, achieve a high channel utilization. The result is

a significant departure from most recent works which only support service differentiation

instead of QoS guarantee.

The studies in Chapter 3 also suggest a good metric channel busyness ratio to represent

the network status, such as throughput, medium access delay and collision probability.

Just as the name implies, channel busyness ratio is a ratio of the time intervals when the

channel is busy due to successful transmissions and collisions to the total time. Based on

the physical carrier sensing and virtual carrier sensing mechanisms of the IEEE 802.11

standard, this metric is very easy to measure and only requires a few simple calculations

at the MAC layer. Hence it can be used to facilitate the regulation of total input traffic to

support QoS.

In Chapter 4, we propose a call admission and rate control scheme to support QoS

guarantee in Wireless LANs. Based on the channel busyness ratio obtained at the MAC

layer, the call admission control algorithm is used to regulate the admission of real-time

and streaming traffic and the rate control algorithm to control the transmission rate of best

effort traffic. As a result, the real-time or streaming traffic is supported with statistical QoS

guarantee, and the best effort traffic can fully utilize the residual channel capacity left by

the real-time and streaming traffic.

In Chapter 5, we further develop the scheme in Chapter 4 into a comprehensive pro-

tocol. Fairness is a major focus of this chapter. We propose a novel three-phase control

mechanism to fairly and efficiently utilize network resource and guarantee a short medium

access delay. The protocol also integrates the three-phase control mechanism with a call









admission control scheme and a packet concatenation scheme into a single unified frame-

work to better support QoS and multiple channel rates besides the efficiency and fairness.

After we examine the performance of wireless LANs and propose a scheme to sup-

port QoS as well as high efficiency, we are wondering whether these techniques can be

applied to multihop case, i.e., MANETs. However, MANETs are much more complicated

than wireless LANs. There are a lot of new challenges, such as the infamous hidden and

exposed terminal problems. Before we come up with any designs, we must first understand

thoroughly what the problems are and how they impact the network performance.

In Chapter 6, we study the impact of physical carrier sensing and virtual carrier sens-

ing mechanisms on the system performance ofMANETs. A theoretical model is developed

to analyze the optimal carrier sensing range to maximize the system throughput when mul-

tiple discrete channel rates coexist in the network. We also study how to utilize the multi-

rate capability of the IEEE 802.11 standard, and which neighbor and channel rate should

be used for each hop transmission. A novel routing metric, bandwidth distance product, is

proposed to perform this task and it can greatly increase the system throughput.

In Chapter 7, we first study the various problems of the medium access control if the

IEEE 802.11 DCF protocol is used, such as the hidden and exposed terminal problems,

receiver blocking problem and intra-flow and inter-flow contention problems. The studies

show that these problems not only impact the efficiency of the MAC protocol but also im-

pact the higher layers' performance, such as unnecessary re-routing activities due to false

route failures and unfairness among multiple flows. Motivated by the analysis of these

problems, we propose a new dual-channel MAC protocol. The new MAC protocol uses

an out-of-band busy tone and two communication channels, one for control frames and the

other for data frames. The newly designed message exchange sequence provides a com-

prehensive solution to all the aforementioned problems. Extended simulations demonstrate

that our scheme provides a much more stable link layer, greatly improves the spatial reuse,

and works well in reducing the packet collisions. It improves the throughput by 8% to 28%









for one-hop flows and by 2~5 times for multihop flows under heavy traffic comparing to

the IEEE 802.11 MAC protocol.

However, sometimes we only have one single channel and one single transceiver. In

this case we need to develop a new efficient MAC protocol other than DUCHA to address

those problems. Therefore, in Chapter 8, we propose a complete single channel solution to

address both hidden and exposed terminal problems. The new solution inserts dummy bits

in the DATA frame and allows the receiver to transmit short busy advertisements during the

transmission time of the dummy bits to notify the hidden terminal of the ongoing transmis-

sion. Because the transmission of DATA frame is protected by the short busy advertisement

signals, we are able to significantly reduce the carrier sensing range to increase the spatial

reuse ratio, which noticeably mitigate the exposed terminal problem. We also demonstrate

that power control in the solution can further remarkably improve the system performance.

In Chapter 9, we study how the physical layer information can be used at the MAC

layer to improve the system performance. We propose a new adaptive packet concatenation

(APC) scheme and demonstrate that APC can improve the system throughput by several

times in both WLANs and MANETs.

In Chapter 10, we focus on the impact of routing metrics on the throughput of se-

lected paths in MANETs. Because MAC layer and Physical layer have a great impact on

the routing algorithm, considering the features of these two layers is a must in a good rout-

ing algorithm. We first perform a comprehensive study on the impact of multiple rates,

interference and packet loss rate together on the maximum end-to-end throughput or path

capacity. A theoretical model is derived to study the path capacity or the maximum end-to-

end throughput of selected paths with consideration of all those factors. We also propose

a new routing metric called interference clique transmission time to efficiently utilize the

information at physical and MAC layers to find good paths. Based on the proposed the-

oretical model, we evaluate the capability of various routing metrics including hop count,

expected transmission times, end-to-end transmission delay or medium time, link rate,









bandwidth distance product, interference clique transmission time, to find a path with high

throughput. The results show that interference clique transmission time is a better routing

metric than all the others.

In Chapter 11, by carefully studying the intra-flow and inter-flow contention problems,

we find that network congestion is closely coupled with the medium access contentions.

Then we propose a framework of distributed flow control and medium access to mitigate

the MAC layer contentions, overcome the congestion and increase the throughput for traffic

flows across shared channel environments. The key idea is based on the observation that, in

the IEEE 802.11 MAC protocol, the maximum throughput for a standard chain topology is

1/4 of the channel bandwidth and its optimum packet scheduling is to allow simultaneous

transmissions at nodes which are four hops away. The proposed fully distributed scheme

generalizes this optimum scheduling to any traffic flow which may encounter intra-flow

and inter-flow contentions. Our scheme has been shown to perform better and achieve

higher throughput at light to heavy traffic load comparing to that when the original IEEE

802.11 MAC protocol is used. Moreover, our scheme also achieves much better and more

stable performance in terms of delay, fairness and scalability with low and stable control

overhead.

The proposed scheme in Chapter 11 provides a good solution of congestion control

at the network and data link layers. However, to support end-to-end reliability required by

various services, such as web traffic and emails, end-to-end flow and congestion control is

also necessary. Chapter 12 studies the close coupling between TCP traffic and medium con-

tention and finds that the TCP sources are very greedy leading to severe network congestion

and medium collisions. And the window based congestion control algorithm becomes too

coarse in its granularity, causing throughput instability and excessively long delay. Based

on the novel use of channel busyness ratio, which we show in Chapter 3 is an accurate sign







7

of the network utilization and congestion status, a new end-to-end congestion control pro-

tocol has been proposed to efficiently and fairly support the transport service in multihop

ad hoc networks.

Finally, Chapter 13 discusses some future research issues including the fairness and

QoS support in MANETs.















CHAPTER 2
PERFORMANCE OF THE IEEE 802.11 DCF PROTOCOL IN WIRELESS LANS

IEEE 802.11 MAC protocol is the de facto standard for wireless LANs, and has also

been implemented in many network simulation packages for wireless multi-hop ad hoc

networks. However, it is well known that, as the number of active stations increases, the

performance of IEEE 802.11 MAC in terms of delay and throughput degrades dramati-

cally, especially when each station's load approaches to its saturation state. To explore

the inherent problems in this protocol, it is important to characterize the probability dis-

tribution of the packet service time at the MAC layer. In this chapter, by modeling the

exponential backoff process as a Markov chain, we can use the signal transfer function of

the generalized state transition diagram to derive an approximate probability distribution of

the MAC layer service time. We then present the discrete probability distribution for MAC

layer packet service time, which is shown to accurately match the simulation data from

network simulations. Based on the probability model for the MAC layer service time, we

can analyze a few performance metrics of the wireless LAN and give better explanation to

the performance degradation in delay and throughput at various traffic loads. Furthermore,

we demonstrate that the exponential distribution is a good approximation model for the

MAC layer service time for the queueing analysis, and the presented queueing models can

accurately match the simulation data obtained from ns-2 when the arrival process at MAC

layer is Poissonian.

2.1 Introduction

The Carrier Sense Multiple Access with Collision Avoidance (CSMA/CA) protocol

used in the IEEE 802.11 MAC protocol has been proposed as the standard protocol for

wireless local area networks (LANs), which has also been widely implemented in many

wireless testbeds and simulation packages for wireless multi-hop ad hoc networks.









However, there are many problems encountered in the higher protocol layers in IEEE

802.11 wireless networks. It has been observed that the packet delay increases dramati-

cally when the number of active stations increases. Packets may be dropped either due to

the buffer overflow or because of serious MAC layer contentions. Such packet losses may

affect high layer networking schemes such as the TCP congestion control and networking

routing maintenance. The routing simulations [19, 108] over mobile ad hoc networks indi-

cate that network capacity is poorly utilized in terms of throughput and packet delay when

the IEEE 802.11 MAC protocol is integrated with routing algorithms. TCP in the wireless

ad hoc networks is unstable and has poor throughput due to TCP's inability to recognize the

difference between the link failure and the congestion. Besides, one TCP connection from

one-hop neighbors may capture the entire bandwidth, leading to the one-hop unfairness

problem [64, 140, 46, 110].

Performance analysis for the IEEE 802.11 MAC protocol could help to discover the

inherent cause of the above problems and may suggest possible solutions. Many papers

on this topic have been published [20, 22, 15, 44, 58, 134, 85]. Cali [20, 22] derived

the protocol capacity of the IEEE 802.11 MAC protocol and presented an adaptive back-

off mechanism to replace the exponential backoff mechanism. Bianchi [15] proposed a

Markov chain model for the binary exponential backoff procedure to analyze and compute

the IEEE 802.11 DCF saturated throughput. All of these papers assume the saturated sce-

nario where all stations always have data to transmit. Based on the saturated throughput

in Bianchi's model, Foh and Zuckerman presented the analysis of the mean packet delay

at different throughput for IEEE 802.11 MAC [44]. Hadzi-Velkov and Spasenovski also

gave an analysis for the throughput and mean packet delay in the saturated case by incor-

porating frame-error rates [58]. Kim and Hou [85] analyzed the protocol capacity of IEEE

802.11 MAC with the assumption that the number of active stations having packets ready

for transmission is large.









To the authors' best knowledge, there is no comprehensive study on the queue dy-

namics of the IEEE 802.11 wireless LANs. The delay analysis is limited to the derivation

of mean value while the higher moments and the probability distribution function of the

delay are untouched. Most of the current papers focused on the performance analysis in

saturated traffic scenarios and the comprehensive performance study under non-saturated

traffic situations is still open.

In this chapter, to address the above issues, we first characterize the probability dis-

tribution of the MAC layer packet service time (i.e., the time interval between the time

instant a packet starts to contend for transmission and the time instant that the packet either

is acknowledged for correct reception by the intended receiver or is dropped). Based on

the probability distribution model of the MAC layer packet service time, we then study

the queueing performance of the wireless LANs at different traffic load based on the IEEE

802.11 MAC protocol. Then, we evaluate the accuracy of the exponential probability distri-

bution model for the MAC layer service time in queueing analysis through both analytical

approach and simulations.

2.2 Preliminaries

2.2.1 Distributed Coordination Function (DCF)

Before we present our analysis for 802.11 MAC, we first briefly describe the main

procedures in the DCF of 802.11 MAC protocol [68]. In the DCF protocol, a station

shall ensure that the medium is idle before attempting to transmit. It selects a random

backoff interval less than or equal to the current contention window (CW) size based on

the uniform distribution, and then decreases the backoff timer by one at each time slot

when the medium is idle (may wait for DIFS followed a successful transmission or EIFS

followed a collision). If the medium is determined to be busy, the station will suspend

its backoff timer until the end of the current transmission. Transmission shall commence

whenever the backoff timer reaches zero. When there are collisions during the transmission

or when the transmission fails, the station invokes the backoff procedure. To begin the










backoff procedure, the contention window size CW, which takes an initial value of CWmin,

doubles its value before it reaches a maximum upper limit CWmax, and remains the value

CWmax when it is reached until it is reset. Then, the station sets its backoff timer to a

random number uniformly distributed over the interval [0, CW) and attempts to retransmit

when the backoff timer reaches zero again. If the maximum transmission failure limit is

reached, the retransmission shall stop, CW shall be reset to CWmin, and the packet shall

be discarded [68]. The RTS/CTS mechanisms and basic access mechanism of IEEE 802.11

are shown in Fig. 2-1.


Idle Period Backoff Successful transmission Backof Collision Backoff
1 1 IoIl e I Ii
DF A C
I I I Collided CTS Timeoul=SIFS+CTS
S T IT F RTS (collided stations)
_______ S S
CW = CW = C CW 2CWmin
CW C i" EIFS=SIFS+ACK+DIFS (collided stations)
(other salons) CW = CWmin
I (other stations)
DIFS DIFS DIFS

Idle Period Backoff Successful transmission Backof a Collision Backoff

I Collided ACK Timeoul=SIFS+ACK
DA F DATA (collided stations)

CW = CWi CW =CW CW = 2CWmin
a lnu EIFS=SIFS+ACK+DIFS (collided stations)
(other stations) CW = CWmin
I (other stations)
DIFS DIFS DIFS


Figure 2-1: RTS/CTS mechanism and basic access mechanism of IEEE 802.11


2.2.2 System Modeling

Each mobile station is modeled as a queueing system, which can be characterized by

the arrival process and the service time distribution. And the saturated status is reached if

each station has heavy traffic and always has packets to transmit. The non-saturated status,

i.e., under light or moderate traffic load, could be characterized by the non-zero probability

that the queue length is zero.

The service time of the queueing system is the MAC layer packet service time defined

in Section 2.1. The IEEE 802.11 MAC adopts the binary exponential backoff mechanism

for the transmission of each packet, which may collide with some other transmissions in









the air at each transmission attempt. And the collision probability pc is determined by the

probability that there is at least one of other stations which will transmit at the same backoff

time slot when the considered station attempts transmission. We assume that this probabil-

ity does not change and is independent during the transmission of each packet regardless of

the number of retransmission suffered. For the saturated case, this approximation has been

used by Bianchi [15] to derive the saturated throughput. And for the non-saturated case,

the collision probability becomes more complex. It depends on the number of stations with

packets ready for transmission and the backoff states of these stations. Between two trans-

mission attempts at the considered station, other stations may complete several successful

transmissions and/or encounter several collisions, and there may be new packet arrivals at

stations no matter whether they are previously contending for transmission or not. Intu-

itively, this approximation becomes more accurate when the number of stations gets larger

for both saturated and non-saturated case. For simplicity, we use the same approximation

for both cases and argue that the collision probability does not change significantly as long

as the input traffic rate from higher layer at each station are still the same during the ser-

vice for each packet. Then we could model the binary exponential backoff mechanism as

a Markov chain and make possible the derivation of the probability distribution of service

time in the next section. Later in this chapter, we will show that the analytical results from

this approximation are consistent with the simulation results very well at the non-saturated

case.

2.3 The Probability Distribution of the MAC Layer Service Time

2.3.1 MAC Layer Service Time

As described in section 2.2, there are three basic processes when the MAC layer trans-

mits a packet: the decrement process of the backoff timer, the successful packet transmis-

sion process that takes a time period of T,,, and the packet collision process that takes

a time period of Too. Here, Tsu, is the random variable representing the period that the









medium is sensed busy because of a successful transmission, and To,, is the random vari-

able representing the period that the medium is sensed busy by each station due to colli-

sions.

The MAC layer service time is the time interval from the time instant that a packet

becomes the head of the queue and starts to contend for transmission to the time instant that

either the packet is acknowledged for a successful transmission or the packet is dropped.

This time is important when we examine the performance of higher protocol layers. Appar-

ently, the distribution of the MAC layer service time is a discrete probability distribution

because the smallest time unit of the backoff timer is a time slot. Ts,, and To,, depend

on the transmission rate, the length of the packet and the overhead (with a discrete unit,

i.e., bit), and the specific transmission scheme (the basic access DATA/ACK scheme or the

RTS/CTS scheme) [15, 68].

2.3.2 Probability Generating Functions (PGF) of MAC Layer Service Time

The MAC layer service time is a non-negative random variable denoted by random

variable Ts, which has a discrete probability of pi for Ts being ti with the unit of one-bit

transmission time or the smallest system clock unit, i=0,1,2,.... The PGF of Ts is given

by

PTs(Z) = iZYt" = poZtO + plZtl + 2Zts2 + ... (2.1)

and completely characterizes the discrete probability distribution of Ts and has a few

important properties as follows:

PT (l) 1
E[Ts] PT,(Z)\ z= P (1) (2.2)

VAR[X] = P (1) + P (1) {P (1)}2


where the prime indicates the derivative.









To derive the PGF of the MAC layer service time, we will model the transmission

process of each packet as a Markov chain in the following subsections. Here we first

discuss how to drive the PGF of the service time from the Markov chain.

The state when the packet leaves the mobile station, i.e., being successfully transmit-

ted or dropped, is the absorption state of the Markov chain for the backoff mechanism. To

obtain the average transition time to the absorption state of the Markov chain, we can use

the matrix geometric approach. However, in the case of Markov Chain for Ts with various

transition times on different branches, it requires a new matrix formulation to accommo-

date different transition times, and its solution always accompanies extraneous complicated

computations [30]. Here, we apply the generalized state transition diagram, from which we

can easily derive the PGF of Ts and obtain arbitrary nth moment of Ts.

In the generalized state transition diagram, we mark the transition time on each branch

along with the transition probability in the state transition diagram (the Markov chain). The

transition time, which is the duration for the state transition to take place, is expressed as

an exponent of Z variable in each branch. Thus, the probability generating function of total

transition time can be obtained from the signal transfer function of the generalized state

transition diagram using the well-known Mason formula [30, 112].

To illustrate how the generalized Markov chain model works, we show one simple

example for a MAC mechanism that allows infinite retransmissions for each packet without

any backoff mechanisms. If the random variable F is defined as the duration of time taken

for a state transition from the state "1" to "2" in Fig. 2-2, its PGF is simply the signal

transfer function of the state transition. In Fig. 2-2, p is the collision probability, 1 p

is the successfully transmitted probability, -Tis the collision time, and 72 is the successful

transmission time. So the PGF of random variable F is

(1 p)Z72
PF(Z) (= )Z2 (2.3)
1 pZT2









This satisfies Equation 2.2, that is, PF(1) = 1 and its mean transition time is


P(1) 1 + 72
1-p


(2.4)


pZI


Figure 2-2: Generalized state transition diagram of one example



On the other hand, we can easily obtain the average collision/retransmission times

Nc, i.e., p/(1 p). Thus the average transition time can be directly obtained as Nc x

71 + 72, which is the same as Equation 2.4.

2.3.3 The Processes of Collision and Successful Transmission

We first study the RTS/CTS mechanisms. As shown in Fig. 2-1, the period of suc-

cessful transmission T,,, equals to


T,,, = RTS + CTS + DATA + ACK + 3SIFS + DIFS


(2.5)


And the period of collision Tol equals to


T,,o = RTS + SIFS + ACK + DIFS = RTS + EIFS


(2.6)


T,,o is a fixed value and its PGF Ct(Z) equals


Ct(Z) ZRTS+EIFS


(2.7)


Tsu, is a random variable determined by the distribution of packet length. In the case

that the length of DATA has a uniform distribution in [1min, Imx], its PGF St(Z) equals


Imax
St (Z) ZRTS+CTS+ACK+3SIFS+DIFS 1 Zi
"max Imin + 1
i lmin


(2.8)









In the case that the length of DATA is a fixed value ID, its PGF St(Z) equals


St (Z) Z= RTS+CTS+D +ACK+3SIFS+DIFS (2.9)

If the basic scheme is adopted, T,,o is determined by the longest one of the collided

packets. When the probability of three or more packets simultaneously colliding is ne-

glected, its probability distribution can be approximated by the following equation,


Pr{Tco i} Pr{/l i,12 <, i} + Pr{/2 1, <, i} Pr{/ ,2 i} (2.10)

where li(i = 1, 2) is the packet length of the ith collided packet. Thus we could obtain that

SImax
C((Z) ZEIFS )2 (2i 21min + 1)Zi (2.11)
(lmax Imin + 1) min

I Imax
St(Z) SIFS+ACK+DIFs 1 i (2.12)
max -mrin +i i(
i=1min
for the case that the length of DATA has a uniform distribution in [1min, Imax], or


Ct(Z)= ZD+EIFS (2.13)


St (Z)= ZID+SIFS+ACK+DIFS (2.14)

for the case that the length of DATA is a fixed value 1D.

2.3.4 Decrement Process of Backoff Timer

In the backoff process, if the medium is idle, the backoff timer will decrease by one for

every idle slot detected. When detecting an ongoing successful transmission, the backoff

timer will be suspended and deferred a time period of Tsuc, while if there are collisions

among the stations, the deferring time will be Tol.

As mentioned in section 2.2, pc is the probability of a collision seen by a packet

being transmitted on the medium. Assuming that there are n stations in the wireless LAN

we are considering and packet arrival processes at all the stations are independent and

identically distributed, we observe that pc is also the probability that there is at least one









packet transmission in the medium among other (n-1) stations in the interference range of

the station under consideration. This yields


Pc 1 [1 (1 po)T)]1"- (2.15)

where po is the probability that there are no packets ready to transmit at the MAC layer in

the wireless station under consideration, and 7 is the packet transmission probability that

the station transmits in a randomly chosen slot time given that the station has packets to

transmit.

Let P,,, be the probability that there is one successful transmission among other (n-1)

stations in the considered slot time given that the current station does not transmit. Then,


Psu n- (1- po)T(1 (1 po)T)(n-2) (n 1)((1 )(n-2)/(n-1) +Pc- 1)
1
(2.16)

Then pc Pc is the probability that there are collisions among other (n-1) stations

(or neighbors).

Thus, the backoff timer has the probability of 1-pc to decrement by 1 after an empty

slot time a, the probability Pc to stay at the original state after Tuc, and the probability of

pc Puc to stay at the original state after Tcoi. So the decrement process of backoff timer

is a Markov process. The signal transfer function of its generalized state transition diagram

is
(1 pe)Z"
Hd(Z) (2.17)
[1 PsucSt[Z) (pC Psuc)Ct[Z)\
From above formula, we observe that Hd(Z) is a function of pc, the number of stations

n and the dummy variable Z.

2.3.5 Markov Chain Model for the Exponential Backoff Procedure

Whenever the backoff timer reaches zero, transmission shall commence. According to

the definition of pc, the station has the probability 1-pc to finish the transmission after Tsuc,

and the probability pc to double contention window size and enter a new backoff procedure









until the maximum retransmission limit is reached after T,,o. Since the decrement process

of backoff timer is a Markov process as discussed above, the whole exponential backoff

procedure is also a Markov process.

Let W be the minimum value of contention window size CWmin plus 1. Following a

similar procedure used by Bianchi [15] and noticing that the transition probability at each

branch of the Markov chain is different from there (which only denoted the value at the

saturated status and did not consider that the contention window is reset after the maximum

a times of retransmissions as defined in the protocols [68], we can obtain (please refer to

Section 2.3.8)



2(1-p.+1)
1-p+ +pcW -i o (2pc)i+W(1-2mp'+1)


where m is the maximum number of the stages allowed in the exponential backoff pro-

cedure (the definition is clarified below). We will use Equations (2.15) and (2.18) in the

queueing analysis to derive the collision probability at different input traffic in Section 2.4.

2.3.6 Generalized State Transition Diagram

Now, it is possible to draw the generalized state transition diagram for the packet

transmission process as shown in Fig. 2-3. In Fig. 2-3, {s(t), b(t)} is the state of the bi-

dimensional discrete-time Markov chain, where b(t) is the stochastic process representing

the backoff timer count for a given station, and s(t) is the stochastic process representing

the backoff stage with values (0, ..., a) for the station at time t. Let m be the "maximum

backoff stage" at which the contention window size takes the maximum value, i.e., CWmax

= 2"(CWmin + 1) 1. At different "backoff stage" i E [0, a], the contention window size










CW, 1 = Wi 1, where Wi = 2i(CWmin + 1) if 0 i < m, and Wi = CWmax + 1 if m < i

< a.


end

Figure 2-3: Generalized state transition diagram for transmission process


As we defined before, the random variable Ts is the duration of time taken for a state

transition from the start state (beginning to be served) to the end state (being transmitted

successfully or discarded after maximum a times retransmission failures). Thus, its Prob-

ability Generating Function (PGF), denoted as B(Z) that is the function of p, n and Z, is




1 The set of CW values shall be sequentially ascending integer power of 2, minus 1,
beginning with CWmin, and continuing up to and including CWmax. [68]









simply the signal transfer function from the start state to the end state given by:


HWi(Z) = o H (Z)/ ) (0 < i m)
HWm(Z), (m < i < a)

H,(Z) = ]^iHWI(Z), 0 i ( a)
j=0
a
B(Z) (1 p)SP(Z) () (pCt(Z))Hi(Z) + (pcCt(Z))a+lHa(Z)
i=0
(2.19)


Since B(Z) can be expanded in power series, i.e.,

oO
B(Z) = Pr(T, i)Z' (2.20)


we can obtain the arbitrary nth moment of Ts by differentiation (hence the mean value and

the variance), where the unit of Ts is slot. For example, the mean is given by

dB(Z)
=E[Ts] d= B z=1 (2.21)
dZ

where p is the MAC layer service rate.

2.3.7 Probability Distribution Modeling

From the probability generation function (PGF) of the MAC layer service time, we

can easily obtain the discrete probability distribution. Fig. 2-4 shows the probability dis-

tribution of the MAC service time at each discrete value. This example uses RTS/CTS

mechanisms. The lengths of RTS/CTS/ACK conform to IEEE 802.11 MAC protocol. Data

packet length is 1000 bytes and data transmission rate is 2 Mbps. The values of the para-

meters are summarized in Table I.

We notice that the envelope of the probability distribution is similar to an exponential

distribution. If we use some continuous distribution to approximate the discrete one, it

will give us great convenience to analyze the queueing characteristics. Fig. 2-4 shows the

approximate probability density distribution (PDF) of Ts and several well-known continu-

ous PDFs including Gamma distribution, log-normal distribution, exponential distribution

















n=10, pc=0.29


n=10, pc=0.001


5 10 15
MAC Service Time (ms)

(c)


20 30
MAC Service Time (ms)

(e)


0.06

0.05

0.04

0.03

0.02

0.01


n=10, pc=0.05














10_J20 30


0 10 20 30
MAC Service Time (ms)

(b)


40 50


0.02


0.015
Li

0.01


0.005


0
0





0.25


0.2


0.15
LL

0.1


0.05


0


100 200 300
MAC Service Time (ms)

(d)


0 5 10 15 20
MAC Serice lime (ms)

(f)


Figure 2-4: Probability distribution of MAC layer service time


0.08

0.07

0.06

S0.05

S0.04

0.03

0.02

0.01


0 100 200 300 4
MAC Service Time(ms)

(a)


0.035

0.03

0.025

S0.02

-0.015

0.01

0 005

0
0







22

Table 2-1: IEEE 802.11 system parameters

Channel Bit Rate 2 Mbit/s
PHY header 192 bits
MAC header 224 bits
Packet payload size 1000Bytes
Length of RTS 160bits + PHY header
Length of CTS 112bits + PHY header
Length of ACK 112bits + PHY header
Initial backoff window size 31
(W)
Maximum backoff stages (m) 5
Short retry limit 7
Long retry limit 4


and Erlang-2 distribution. We observe that the log-normal distribution provides a good

approximation for almost all cases (not only for cases at the high collision probability but

also for cases at the low collision probability), and also has a very close tail distribution

match with that of Ts. In addition, the exponential distribution seems to provide a rea-

sonably good approximation except for cases at very low collision probability, where it is

more like a deterministic distribution. Here, the PDF of Ts is obtained by assuming that

the probability density function is uniform in a very small interval and is represented by a

histogram while other continuous PDF is determined by the value of mean and/or variance

of Ts. Here, we use 5 ms as the interval in the histogram because the distribution of the

delay concentrates around the integer times of the successful transmission period for each

packet which approximates 5 ms for packets with 1000 bytes long.

We also notice that pc has different saturation values for different n. If the mobile

station always has packets to transmit, i.e., in the saturation state, the idle probability po

takes the minimum value 0. So, according to formulae (2.15) and (2.18), we can obtain the

saturation value of pc by setting po as 0 in Table II.

Table 2-2: Saturation value of collision probability

n 5 9 17 33 65
Max pc 0.1781 0.2727 0.3739 0.4730 0.5692








23

0.03 pc=0.20
i -- n=9
0.025 n=17
---- n=33
0.02

00.015
0
0-
0.01

0.005

Ol
0 100 150 200
MAC Layer Service Time (ms)

Figure 2-5: PDF of service time


Fig. 2-5 shows the distribution of Ts at different number of mobile stations, which

mainly depends on pc and hardly depends on n. Fig. 2-6 shows the mean value of Ts at

different collision probability. The maximum of Ts for different n, which is reached when

pc takes the saturation value, is marked. We observe that the distribution of Ts mainly

depends on pc and is determined by the number of the active stations at saturation status

when pc reaches the saturation value. We will discuss how to obtain the value of pc at

different traffic load in the following section.


o30o
--n=5-,,,,,,,,'

-250 n=6
E n=9
-200 -- n=17
----- n=33
0150
o '


I 50

0
0 0.1 0.2 0 0.4
Collision Probabiity pc

Figure 2-6: Mean of service time


2.3.8 Derivation of Transmission Probability

This section derives the transmission probability r, i.e., the packet transmission prob-

ability that the station transmits in a randomly chosen slot time given that it has packets to









transmit. We follow the similar notations in the paper [15]. {s(t), b(t)} and Wi have been

defined in section 2.4.6. F. Let P{ii, k io, ko} be the short notation of one-step transition

probability and P{il, kllio, ko} = Pr{s(t + 1) = i, b(t + 1) = kl|s(t) = io,b(t) = ko}.

Then the only non null one-step transition probabilities are

P{i, ki,k+1} =1 k [0,Wi 2] i [0, c]

P{0, ki,0} = (1 p)/Wo k [0, Wo 1] i [0, (2.22)
(2.22)
P{i, kli-,0}= pc/W k c [0, Wi-1] i [1,a]

P{0, ka, 0}= 1/Wo k c [0,Wo 1]

These equations account for the facts that: the backoff timer is decremented; the back-

off timer starts from stage 0 after a successful transmission; the backoff timer starts from a

new stage after an unsuccessful transmission; the contention window size is reset and the

backoff timer starts from stage 0 when the maximum transmission failure limit is reached,

respectively.

Let bi,k lim Pr{s(t) = i,b(t) = k},0 i < a, 0 < k < Wi be the stationary
t--oo
distribution of the Markov chain. First, note that


bi-1,o pc = bi,o bi,o = p'bo,o 0 < i < a (2.23)

and
W1I-k bo,o + (1 p) j bj,o i =0
bi,k -= x 0 (2.24)
Pc bi-l,o 0 < i < a
By means of equation (2.23), equation (2.24) can be simplified as


bi,k k bo (0 < i (< a, 0 < k < Wi 1) (2.25)
Wi










Thus, bo,o can be finally determined by imposing the normalization condition, which

simplifies as follows:

a Wi-1 a W-1 a a
1 O bik bi,o k bi,o l boo Yi 2 WPc
i=0 k=0 i=0 k=0 i=0 i=0

PCOi 2W a bo,o i=0
2 m-1 a

i=0 i=m

As any transmission occurs when the backoff time counter equals zero, regardless of

the backoff stage, the probability r that a station, which has packets to transmit, transmits

in a randomly chosen slot time is

a 1 a+l
7- bo 1 -- p bo,o (2.27)
t1 PC
i=0

which can be simplified as



1-p++(1-pc)W(E o (2pc)i)
2(1-p~+1)
1-p+ +pcW E-' (2pc)i+W(1-22mp+1)
a > rn


2.4 Queueing Modeling and Analysis

2.4.1 Problem formulation

Many applications are sensitive to end-to-end delay and queue characteristics such as

average queue length, waiting time, queue blocking probability, service time, and goodput.

Thus, it is necessary to investigate the queueing modeling and analysis for wireless LANs

to obtain such performance metrics.

A queue model can be characterized by the arrival process and the service time distri-

bution with certain service discipline. We have characterized the MAC layer service time

distribution in the previous section. In this chapter, we assume that the packet arrivals at

each mobile station follow the Poisson process or a deterministic distribution with average

arrival rate A. The packet transmission process at each station can be modeled as a general









single "server". The buffer size at each station is K. Thus, the queueing model for each

station can be modeled as an M/G/1/K when Poisson arrivals of packets are assumed.

2.4.2 The steady-state probability of the M/G/1/K queue

Let p, represent the steady-state probability of n packets in the queueing system, and

let 7rrepresent the probability of n packets in the queueing system upon a departure at the

steady state, and let P {pjj} represent the queue transition probability matrix:


Pij = Pr{XI = jlX, = i} (2.29)

where X, denotes the number of packets seen upon the nth departure.

To obtain pij, we define


k, Pr{narrivals during service time Ti} = n Pr{Ts = i}
i=0
where A is the average arrival rate. We can easily obtain

ko ki k2 .. kK-2 t l- -2 k,

ko ki k2 ... kK-2 1 K-2 kn

P= {Pi}= 0 ko k1i .. kK-3 1- Z03k (2.30)


0 0 0 ... ko 1 ko

Moreover, we notice that


ko = B(e-), k (2.31)
(-1)"n! aiA"

where B(e-A) is obtained by replacing Z with e- in equation (2.19), i.e., the PGF of the

MAC layer service time Ts.

According to the balance equation:


7rP =


(2.32)









where 7r = {7, } and the normalization equation, we can compute the 7r. For the finite

system size K with Poisson input, we have [53]

T0o 7n 1
P0o Pn- (0 7io + P 7op + +

where p is the traffic intensity and p = AE[Ts].

If we can approximate the distribution of MAC service time by an exponential distri-

bution, the steady-state probability for the M/M/1/K model [53] is given by:

K
Po = [i0opi P (p)ipo, (0 < i < K) (2.34)


2.4.3 Conditional Collision Probability pc and Distribution of MAC Layer Service
Time

From above derivation, we know that po is a function of pc, A, and n. So we can

compute pc under different values of A and n with the help of (2.15) and (2.18) using some

recursive algorithm. Thus, we can obtain the distribution of MAC service time at different

offered load according to the results obtained in section 2.4.6. Here we assume that the

packet arrival process at each station is independent and identical distributed, and hence

we could obtain the aggregate performance of wireless LAN from the queueing analysis in

this section.

2.4.4 Performance Metrics of the Queueing System

The average queue length, blocking probability, and average waiting time including

MAC service time are given by
K 1 L
L i p = Pk 1 W (2.35)
i=o + A(1 PB)

2.4.5 Throughput

If we know the blocking probability pB, then the throughput S at each station can be

computed easily by


S= (1 )(1 -pj1)


(2.36)











where p1+ is the packet discard probability due to transmission failures.

2.4.6 Numerical Results

Fig. 2-7 shows the results for the major performance metrics. All of them have

a dramatic change around the traffic load of 1.1-1.5 Mbits/sec. This is because that the

collisions increase significantly around this traffic load, resulting in much longer MAC

service time for each packet.

0.5 50
04- --------- -----------4-n=3 n=33
( -~ r i-y s if f _-e- n=17

S10
04 -----------1



0.2 ~---------i--- F -i
o3/~ ~ ~ ~ ~ ~~C 20 -----------------....... ------ n -- r.' ..... -- I i------n------

0.... --........

Oo0--- 0.5 1 15 2 O5--
U 0
0 05 1 15 2 0 0.5 1 15 2 0 05 1 1
Offered Load (Mbits/sec) Offered Load (Mbits/sec) Offered Load (Mbits/sec)
08.


S 0 05 1 15 2
Offered Load (Mbits/sec)

Figure 2-7: Queue characteristic


x 103
3
-4-- n=33
S< -e- n=17
1 2 1 -- n=9 ----.. ------
v ,. -- n=5
I - - -

o 6...---- ........ ..
0 0.5 1 15 2


Offered Load (Mbits/sec)


s


From the results, we observe that all the metrics are dependent on the collision prob-

ability pc. Fig. 2-7 shows that pc mainly depends on the total traffic in the non-saturated

scenario. On the other hand, pc is affected by the total number of packets attempting to

transmit by all neighboring stations. In the non-saturated case, when all arriving packets

are immediately served by the MAC layer, the queue length may reach zero and the cor-

responding station will not compete for the medium. However, in the saturated scenario,

i.e., the stations always have packets to transmit, the total number of packets attempting to

transmit equals to the total number of neighboring stations, hence pc is mainly dependent

on the total number of neighboring stations as we expect.


S0.6
2
g0.4
0

0


2









The MAC layer service time shows similar change at different offered load, because

it is dependent on the pc. All other performance metrics are dependent on the distribution

of the MAC layer service time, so they show the similar change in the figures. The average

queue length is almost zero at the non-saturated state and reaches almost maximum length

at the saturated state. The average waiting time for each packet in the queue almost equals

to zero at the non-saturated state and reaches several seconds at the saturated state. The

queue blocking probability is zero at the non-saturated state when the traffic load is low,

and linearly increases with the offered load at the saturated state. The throughput linearly

increases with the offered load at the non-saturated state and maintains a constant value

with different total number of transmitting stations at the saturated state. The throughput at

saturated state decreases when the number of stations increases because collision probabil-

ity climbs up with the number of stations. This is consistent with the results of saturation

throughput found by Bianchi [15] where the author indicates that the saturated throughput

decreases as n increases under a small initial size of the backoff window given a specific

set of system parameters. In addition, the packet discarding probability at MAC layer is

much smaller than the queue blocking probability.

In summary, all these results indicate that IEEE 802.11 MAC works well in the non-

saturated state at low traffic load while its performance dramatically degrades at the sat-

urated state, especially for the delay metric. Besides, at the non-saturated state, the per-

formance is dependent on the total traffic and indifferent to the number of transmitting

stations. At the saturated state, the number of transmitting stations is much more important

to the whole performance. The similar phenomena have been observed for the distribution

of MAC service time shown in section.

2.5 Performance Evaluation

2.5.1 Simulation Environments

In our simulation study, we use the ns-2 package [41]. The wireless channel capacity is

set to 2Mbps. Data packet length is 1000 bytes, and the maximum queue length is 50. The











radio propagation model is Two-Ray Ground model. We use different numbers of mobile

stations in a rectangular grid with dimension 150m x 150m to simulate the Wireless LAN.

All stations have the same rate of packet inputs. The MAC protocol uses the RTS/CTS

based 802.11 MAC and other parameters are summarized in Table I.

2.5.2 Probability Distribution of MAC Layer Service Time

Fig. 2-8 shows the simulation results of the MAC layer service time in the network

with 17 mobile stations and total traffic of 0.2, 0.8 and 1.6 Mbps, respectively. It displays

good match on the probability density functions between the analytical result and that from

simulation. Notice that, similarly with Fig. 2-4, the PDFs shown in Fig. 2-8 are histogram

approximations of the discrete probability distribution obtained from both analysis and

simulations.

Offered Load 800Kbitslsec among 17 stations
Offered Load 200Kbitssec among 17 stations 0.14 io Offered Load 1600Kbitslsec among 17 stations
0.2 -- Simulation 0.02
-*- Simulation 0.12 A~nalysis Simulation
--e- Analysis -- Analysis
015 01 0.015
-008
0 1 -0.06 0.01

0.04
0.05 0.005

0 5 10 15 20 25 0 10 20 30 40 50 0 100 200 300 400 500
MAC Layer Sevice Time (ms) MAC Layer Service Time (ms) MAC Layer Servce Time (ms)

Figure 2-8: MAC layer packet service time



Our results indicate the distribution of MAC layer service time is independent of the

packet input distribution whether it is deterministic or Poisson distributed. It mainly de-

pends on the total traffic in the network before saturation and on the number of mobile

stations after saturation, which is consistent with the analysis.

2.5.3 Comparison of M/G/1/K and M/M/1/K Approximations with Simulation Re-
sults

Exponential distribution is a memoryless distribution. If we can model the MAC layer

service time as this distribution, it will give us great convenience to predict the system










performance, such as throughput, link delay, packet discarding ratio. The problem is how

good this approximation is for our modeling.

As we said in section 2.4, the exponential distribution seems to be a good approxima-

tion for the MAC layer service time. In Fig. 2-9 and 2-10, we compare it with the derived

discrete probability distribution in the queueing analysis to check its goodness to predict

the MAC throughput, packet waiting time, queue blocking probability and average queue

length. Here, we assume that the system has Poisson arrivals. We use two queueing models

for these two distributions: M/M/1/K and M/G/1/K. Fig. 2-9 and 2-10 show the results for

the WLAN with 9 mobile stations.


S0.3
a

S0.2
0
m

D 0.1
0


(c) (d)

Figure 2-9: Comparisons between M/G/1/K, M/M/1/K models and simulation





















0.4 0.6 0.8
Offer Load (Mbits/sec)


Figure 2-10: Average waiting time in non-saturated status


From Fig. 2-9 and 2-10, we observe that M/M/1/K model give a close approximation

to the M/G/1/K model for some performance metrics. Both models have almost the same

throughput and queue blocking probability. However, when the mobile stations are at the

saturated state, M/M/1/K gives a large prediction error for the average queue length and

average waiting time, and the difference is small except at the turning point between non-

saturated state and the saturated state, where a dramatic change of the system performance

is shown. The M/G/1/K model always provides better approximation for all performance

metrics.

We also compare the results of queueing models with the simulation in Fig. 2-9

and 2-10. Two queueing models show very close approximations with the simulation

results for all performance metrics when mobile stations are in the non-saturated state.

However, there are distinct differences between them when the system is in the saturation

state. This is because that the Markov chain model overestimates the average MAC layer

service time about 10 % in the saturation state compared to the simulation results from

ns-2, as showed in Fig.2-11. The reasons may be that the Markov chain model does not

capture all the protocol details and/or the implementation considerations of IEEE 802.11

MAC protocols in ns-2. Thus, the simulation results have higher throughput, lower queue

blocking probability, smaller average queue length and smaller average waiting time at

saturated state.










60
-- MIG/1/K
S50 -- M/M/1/K
E --- simulation
E 40
I-I
*I 30 ------------------- ----------------- ----------------- --------------- ----------
S30
U)
8 20

< 10

0
0 0.5 1 1.5 2
Offered Load (Mbits/sec)

Figure 2-11: Average MAC layer service time


With extensive simulations for different number of mobile stations in randomly gen-

erated wireless LANs, we have concluded that the Markov chain models seem to always

give an upper bound of the average MAC layer service time. Thus, the queueing models

using the distribution of the service time give a lower bound of the throughput, and upper

bounds of queueing blocking probability, average queue length and average waiting time

compared with simulations of ns-2. Therefore, our analytical models can always be useful

to come up with the performance estimates for design purpose.

2.6 Conclusions

In this chapter, we have derived the probability distribution of the MAC layer service

time. To obtain this distribution, we use the signal transfer function of generalized state

transition diagram and expand the Markov chain model to the more general case for the ex-

ponential backoff procedure in IEEE 802.11 MAC protocols. Accurate discrete probability

distribution and approximate continuous probability distributions are obtained in this chap-

ter. Based on the distribution of the MAC layer service time, we come up with a queueing

model and evaluate the performance of the IEEE 802.11 MAC protocol in Wireless LANs

in terms of throughput, delay, and other queue performance metrics. Our results show that

at the non-saturated state, the performance is dependent on the total traffic and indifferent









to the number of transmitting stations, and at saturated state, the number of transmitting

stations affects the performance more significantly.

In addition, the analytical results indicate that exponential distribution may provide a

good approximation for the MAC layer service time in the queueing analysis. The queueing

models discussed in this chapter can accurately estimate various performance metrics of

WLAN in the non-saturated state which is the desired state for some application with a

certain QoS requirement because there is no excessive queueing delay as that in saturated

state. And for WLANs in the saturated state, the queueing models give a lower bound for

the throughput, and upper bounds for queueing blocking probability, average queue length

and average waiting time compared with simulation results obtained from ns-2.















CHAPTER 3
HOW WELL CAN THE IEEE 802.11 DCF PROTOCOL SUPPORT QOS IN
WIRELESS LANS

This chapter studies an important problem in the IEEE 802.11 DCF based wireless

LAN: how well can the network support quality of service (QoS). Specifically, we analyze

the network's performance in terms of maximum protocol capacity or throughput, delay,

and packet loss rate. Although the performance of the 802.11 protocol, such as through-

put or delay, has been extensively studied in the saturated case, we demonstrate that the

maximum protocol capacity can only be achieved in the non-saturated case, and is almost

independent of the number of active nodes. By analyzing the packet delay, consisting of

the MAC service time and waiting time, we derive accurate estimates for delay and delay

variation when the throughput increases from zero to the maximum value. Packet loss rate

is also given for the non-saturated case. Furthermore, we show that the channel busyness

ratio provides precise and robust information about the current network status, which can

be utilized to facilitate QoS provisioning. We have conducted a comprehensive simulation

study to verify our analytical results and to tune the 802.11 to work at the optimal point

with the maximum throughput and low delay and packet loss rate. The simulation results

show that by controlling the total traffic rate, the original 802.11 protocol can support strict

QoS requirements, such as those required by voice over IP or streaming video, and at the

same time, achieve a high channel utilization.

3.1 Introduction

Because of its simple deployment and low cost, the IEEE 802.11 wireless LAN [68]

has been widely used in recent years. It contains two access methods, i.e., Distributed

Coordination Function (DCF) and Point Coordination Function (PCF), with the former

being specified as the fundamental access method. Despite its popular use, currently only










Table 3-1: QoS requirements for multimedia services
Class Application One-way transmission Delay Packet loss
delay variation rate
Real-time VoIP, videoconferencing <150ms(preferred), 1ms* 1%(video),
<400ms(limit) .;' (audio)
Streaming Streaming audio and video up to 10s lms* 1%
Best effort E-mail, file transfer, web minutes or hours N/A Zero
browsing
Playout buffer (or jitter buffer) can be used to compensate for delay variation


best effort traffic is supported in DCF. Section 3.2 describes the 802.11 protocol in more

detail.

For the IEEE 802.11 wireless LAN to continue to thrive and evolve as a viable wireless

access to the Internet, quality of service (QoS) provisioning for multimedia services is

crucial. As shown in Table 3-1, for real-time, streaming, and non-real-time (or best effort)

traffic, the major QoS metrics include bandwidth, delay, delay jitter, and packet loss rate

[73, 74]. Guaranteeing QoS for multimedia traffic, however, is not an easy task given that

the 802.11 DCF is in nature contention-based and distributed, and hence renders effective

and efficient control very difficult. In addition, other problems such as hidden terminals

or channel fading make things worse. To address these challenges, current research works

([1, 125, 161, 153] and references therein) and the enhanced DCF (EDCF) defined in the

IEEE 802.1 le draft [72, 31] tend to provide differentiated service rather than stringent QoS

assurance.

However, we have not yet well understood the question of how well the IEEE 802.11

WLAN can support QoS when many researchers start to believe that service differentiation

is the best that the 802.11 can achieve. In this chapter, we endeavor to address this problem

through both theoretical analysis (Section 3.3) and simulations (Section 3.4).

We develop an analytical model to assess the capability of the 802.11 for supporting

major QoS metrics, i.e., throughput, delay and delay variation, and packet loss rate. While

current literature on performance analysis is focused on the derivation of throughput or

delay in the saturated case, we find that the optimal operating point for the 802.11 to work









at lies in the non-saturated case.1 At this point, we analytically show that the maximum

throughput is achieved almost independent of the number of active nodes, and the delay

and delay variation is low enough to satisfy stringent QoS requirements of the real-time

traffic. Thus the 802.11 WLAN can perform very well in supporting QoS, as long as it is

tuned to the optimal point. Since an accurate indicator of the network status is essential

to effective tuning, we also demonstrate that the channel busyness ratio, which is easy to

obtain and accurately and timely represents the network utilization, can be used to design

schemes such as call admission control or rate control in the WLAN. We will present such

schemes in a subsequent chapter.

In Section 3.5, we show that our analytical results are still valid even when the effect

of channel fading is taken into account. Also, we discuss the possible implications arising

due to the employment of a prioritized 802.11 DCF. Finally, Section 3.6 concludes this

chapter.

3.2 Preliminaries

3.2.1 Operations of the IEEE 802.11

The basic access method in the IEEE 802.11 MAC protocol is DCF (Distributed Coor-

dination Function), which is based on carrier sense multiple access with collision avoidance

(CSMA/CA). Before starting a transmission, each node performs a backoff procedure, with

the backoff timer uniformly chosen from [0, CW-1] in terms of time slots, where CW is

the current contention window. When the backoff timer reaches zero, the node transmits a

DATA packet. If the receiver successfully receives the packet, it acknowledges the packet

by sending an acknowledgment (ACK). If no acknowledgment is received within a speci-

fied period, the packet is considered lost; so the transmitter will double the size of CW and

choose a new backoff timer, and start the above process again. When the transmission of a



1 Note that a similar fact has been observed for the Aloha or Slotted Aloha, where the
maximum throughput is achieved only when traffic arrives at a certain rate [11].









packet fails for a maximum number of times, the packet is dropped. To avoid collisions of

long packets, the short RTS/CTS (request to send/clear to send) frames can be employed.

Note that the IEEE 802.11 MAC also incorporates an optional access method called

PCF (Point Coordination Function), which is only usable in infrastructure network config-

urations and is not supported in most current wireless cards. In addition, it may result in

poor performance as shown in previous research [126, 145]. In this chapter, we thus focus

on the 802.11 DCF.

3.2.2 Related Work

To date two threads of research have examined the property or performance of the

IEEE 802.11: performance analysis, and performance and/or QoS enhancements.

Performance Analysis: The first thread was devoted to building analytical models to

characterize the behavior of the 802.11, and deriving the protocol capacity or delay perfor-

mance [15, 21, 44, 58, 66, 134, 154, 160]. Bianchi [15] proposed a Markov chain model

for the binary exponential backoff procedure. By assuming the collision probability of

each node's transmission is constant and independent of the number of retransmissions, he

derived the saturated throughput for the IEEE 802.11 DCF. Based on the saturated through-

put derived in Bianchi's model, Foh and Zuckerman [44] used a Markovian state dependent

single server queue to analyze the throughput and mean packet delay. Cali et al. [21] stud-

ied the 802.11 protocol capacity by using a p-persistent backoff strategy to approximate

the original backoff in the protocol. Again, the focus is on the saturated throughput. In ad-

dition to collisions, Hadzi-Velkov and Spasenovski took the effect of frame error rate into

account in their analysis of saturated throughput and delay [58]. We derived an approxi-

mate probability distribution of the service time, and based on the distribution, analyzed

the throughput and average delay [154, 160]. As noticed, most works were focused on the

analysis of throughput and delay in the saturated case. Moreover, none of these systemati-

cally considered the delay and delay variation in the non-saturated case, let alone obtained

accurate estimates for them.









Performance and/or QoS Enhancements: The second thread of the research on

the 802.11 DCF explored various ways to improve throughput [13, 20, 85, 90] or provide

prioritized service, namely, service differentiation [1, 81, 114, 119, 125, 137].

Based on the work [21], Cali et al. attempted to approach the protocol capacity by

replacing the exponential backoff mechanism with an adaptive one [20]. Kim and Hou

developed a model-based frame scheduling algorithm to improve the protocol capacity of

the 802.11 [85]. Two fast collision resolution schemes were proposed by Bharghavan [13]

and Kwon et al. [90], respectively. The idea is to use two channels or to adjust backoff

algorithms to avoid collisions, thereby providing higher channel utilization.

To provide service differentiation, Ada and Castelluccia [1] proposed to scale the con-

tention window and use different inter frame spacing or maximum frame length for ser-

vices of different priorities. Two mechanisms [125], i.e., virtual MAC and virtual source,

were proposed to enable each node to provide differentiated services for voice, video, and

data. By splitting the transmission period into a real-time one and a non-real-time one,

the real-time traffic is supported with QoS guarantee [114]. However, the DCF mode was

dramatically changed. The Blackbust [119] provided high priority for the real-time traf-

fic. Unfortunately, it imposes special requirements on high priority traffic and is not fully

compatible with the existing 802.11 standard. In summary, if the semantics of the 802.11

DCF is maintained, all the works mentioned above can only support service differentiation.



Our studies can be considered to be a convergence between these two threads of re-

search; however, it improves on both sides. We thoroughly study the QoS performance

of the 802.11 in terms of throughput, delay and delay variation, and packet loss rate.

Moreover, we discover the optimal operating point at which, in addition to achieving the

theoretical maximum throughput, the 802.11 WLAN is capable of supporting strict QoS

requirements for the real-time traffic, rather than only providing prioritized service.









3.3 Analytical Study of the IEEE 802.11

This section focuses on the analysis of the performance of the IEEE 802.11 DCF. Note

that in the following analysis, the hidden terminal problem is ignored. This is because in

a typical wireless LAN environment, every node can sense all the others' transmissions,

although it may not necessarily be able to correctly receive the packets from all other

nodes.

3.3.1 Maximum Throughput and Available Bandwidth

To simplify the analysis and yet reveal the characteristics of the IEEE 802.11 MAC

protocol, we assume that the traffic is uniformly distributed among the nodes. The total

number of nodes is n. The transmission probability for each node in any time slot is pt.

Note that here a time slot at the MAC layer could be an empty backoff time slot, a period

associated with a successful transmission, or a period associated with a collision [15, 68].

Obviously, we obtain the following equations:


pi = (1 pt)"

ps = npt(l pt)n- (3.1)

Pc = 1 Pi Ps

where pi is the probability that the observed backoff time slot is idle, ps is the probability

that there is one successful transmission, and pc is the collision probability that there are at

least two concurrent transmissions at the same backoff time slot. If we define Tsu, as the

average time period associated with one successful transmission, and Tooi as the average

time period associated with collisions, we know [68]

Tsu, = rts + cts + data + ack + 3sifs + difs
(3.2)
Too = rts + sifs + cts + difs = rts + eifs

for the case where the RTS/CTS mechanism is used, and

Tsc = data + ack + sifs + difs
Sdata* + acktieout + di.3)
Tcoi = data* + acktimeout + difs










RTS/CTS scheme with different number of nodes

n= 5
0.8 n=10 channel busyness ratio'
n=300 --- ^^ \



0 .4 ,/ ,/

0.2


104 10-3 10-2 10-1 100
collision probability p

Figure 3-1: Channel busyness ratio and utilization


for the case where there is no RTS/CTS mechanism, where data and data* (please re-

fer to [15] for derivation of data*) are the average length, in seconds, for the successful

transmission and collision of the data packets, respectively. Thus, it can easily obtained

that
.Ri pio
Pi,++PsTsuc+PcTcol

Rb 1 Ri (3.4)

s Picr+psTsuc+PcTcol
where a is the length of an empty backoff time slot, Ri is the channel idleness ratio, Rb

is the channel busyness ratio, and R, is the channel utilization. Once we obtain R,, the

normalized throughput s is expressed as


s = Rs x datalT,,,su, (3.5)


and the absolute throughput is s times the bit rate for data packets.

In most cases, we are more interested in the packet collision probability p observed at

each individual node, since it can be used to calculate QoS metrics for the traffic traversing

the node. In other words, p is the probability that one node encounters collisions when

it transmits. Also, p is the probability that there is at least one transmission among the







42

Table 3-2: IEEE 802.11 system parameters

Bit rate for DATA packets 2 Mbps
Bit rate for RTS/CTS/ACK 1 Mbps
PLCP Data rate 1 Mbps
Backoff Slot Time 20 ps
SIFS 10 ps
DIFS 50 ps
Phy header 192 bits
MAC header 224 bits
DATA packet 8000 bits + Phy header
+ MAC header
RTS 160 bits + Phy header
CTS, ACK 112 bits + Phy header


neighbors in the observed backoff time slot. We thus link p to pt as


p = 1 (1 pt)'-1 (3.6)


It can be seen that the collision probability increases with the increase in the number of

neighboring nodes or in the traffic at each of these nodes. In this sense, p reflects the

information about both the number of neighboring nodes and the traffic distribution at

these nodes.

According to the above equations, we can express Rb, Rs, and s as a function of p,

which are shown in Fig. 3-1. All the parameters involved are indicated in Table 3-2 and

most are the default values in the IEEE 802.11. In Fig. 3-1, three cases, i.e., n = 5, 10, and

300, are considered. It is important to note that for each specific n, there exists a maximum

value of p, denoted by MAX(p), at which the network operates in the saturated status,

i.e., each of the n nodes always has packets in the queue and thus keeps contending for the

channel. Based on the works [15, 154, 160], we know that in saturated status, the larger the

number of nodes, the greater the collision probability. More precisely, MAX(p) = 0.105,

0.178, 0.290, 0.546, 0.701, 0.848 for n = 3, 5, 10, 50, 128, 300, respectively. Next we

present some important observations from Fig. 3-1.









Channel busyness ratio: an accurate sign of the network utilization

First, we find that the channel busyness ratio is an injective function of the collision

probability. In fact, this can easily be proved. Second, when p < 0.1, Rb is almost the same

as Rs, namely

R, a Rb. (3.7)

This is not hard to understand. When the collision probability p is very small, the chan-

nel resource wasted in collisions is so minor that it can be ignored. Third, the normalized

throughput almost stays unchanged when p increases from 0.1 to 0.2, although it reaches

the maximum value around p = 0.2. Finally, the maximum throughput is almost insensitive

to the number of active nodes. Given these observations and the fact that the throughput

is proportional to Rs, we therefore could use the measured channel busyness ratio Rb to

accurately estimate the throughput from zero to the maximum value. Note that this is very

simple and useful to each node: it can monitor the throughput of the whole WLAN by sim-

ply measuring the channel busyness ratio, which can be easily done since the IEEE 802.11

is a CSMA-based MAC protocol, working on the physical and virtual carrier sensing mech-

anisms. On the other hand, when Rb exceeds a certain threshold thb, severe collisions can

be observed in the WLAN.

Maximum throughput

Fig. 3-1 also shows that the throughput begins to decrease when p is greater than a

certain value, and could decrease to zero when p becomes very large. To ensure that the

network is always working with a high throughput, it is important for us to find the critical

turning point, i.e., when the IEEE 802.11 will achieve the maximum throughput, and how

the maximum throughput depends on network characteristics such as the number of node

n and traffic.









Combining Equation (3.1)(3.4)(3.6), we can write Rs as a function of p. To obtain the

maximum throughput, we take the derivative of R, with respect to p and let it equal 0:


dp

Meanwhile, we know that p is upper bounded by MAX(p). Therefore, if proot is the

root of Equation (3.8), we obtain the value of p, denoted by p*, with which the maximum

throughput is achieved:

p* MIN(proot, MAX(p)), (3.9)

By applying p* to Equation (3.5), we get the maximum normalized throughput of the IEEE

802.11 at different n, as shown in Fig. 3-2(a) and 3-2(b). Here two important points are

noted.

Maximum throughput is achieved in the non-saturated case, rather than in the

saturated case when n > 5. This is the very reason that we argue the network should

work in non-saturated case. When n > 5, the normalized throughput arrives at the maxi-

mum value around p = 0.196, much smaller than the collision probability in the saturated

status, i.e., MAX(p), as clearly seen in Fig. 3-2(a). p = 0.196 means there are 5 or 6

nodes simultaneously contending for the channel, which can be derived from the inverse

function of MAX(p) as shown earlier. In addition, the maximum throughput achieved is

not sensitive to the number of nodes, n. It is rather stable as n increases.

Maximum throughput can be achieved by controlling the total input traffic rate

if no modification to the MAC protocol is allowed. As revealed in Fig. 3-2(b), rather

than letting p = p* for each n, if we simply let p < 0.1 or p < 0.05, the achieved normal-

ized throughput only drops by 0 '".'. and 4.' ,, respectively, compared to the maximum

normalized throughput. This is a very nice and important feature in the sense that as long

as each node in the network can keep the collision probability p below a certain value, say

0.1, instead of p*, which is dependent on n, the maximum throughput is well approached.

Thus, by maintaining a small collision probability in the Wireless LAN, which can be done












0.8

0.6
Proot
0.4 // MAX(p)


0
0 2 7 -- -- X ---- X ---- ----- ---- ---- ?


0 50 100 150 200 250 300
Number of nodes n
(a) Proot and the collision probability in the satu-
rated status

0.72 --
0.7
0.68
0.66
0.64 P=P
p<=0.1
0.62 p<=0.05
0.6 p=MAX(p)
0 50 100 150 200 250 300
Number of nodes n
(b) Maximum normalized throughput with different con-
straints on collision probability p

Figure 3-2: Collision probability and maximum normalized throughput with RTS/CTS and
payload size of 8000bits


through controlling the total input traffic rate, we can achieve high throughput. This in fact

is consistent with our observation in Fig. 3-1, where Rb / Rs when p < 0.1.

Note that in addition to achieving high throughput, keeping a small collision proba-

bility helps reduce delay. Since the time wasted due to collision could be neglected, the

contention delay is very small, which is crucial in providing low delay for the real-time

traffic and will be discussed in detail in section 3.3.2.

Available bandwidth

The total available bandwidth BW, of the wireless LAN, or the available traffic rate

the network could further accommodate, can be easily obtained by subtracting the current

throughput from the maximum throughput.









Although it is not easy for each individual node to know the current total throughput

if it does not decode everything received, the node can be aware of the available bandwidth

by virtue of the channel busyness ratio, which could be easily acquired as described earlier.

Especially, when p < 0.1, Rb / Rs. Thus BW, can be calculated as follows:

W BW(thb Rb)data/T,,, (thb > Rb)

0 ,(thb < Rb)

where BW is the transmission rate in bits/s for the DATA packets, and thb is a threshold of

Rb and proportional to the maximum throughput.

Impact of payload size and the RTS/CTS mechanism

Thus far we have conducted our analysis without considering the impact of payload

size and the RTS/CTS scheme on the throughput. In this subsection we study this impact.

Fig. 3-3 presents the analytical results, where RTS/CTS is used or not used, and various

payload sizes are considered.

We find no matter whether RTS/CTS is used, the throughput increases along with

the payload size. But this is not necessarily true for channel utilization, such as the case

that RTS/CTS is not used in the saturated case. The reason is the following. In the sat-

urated case, given n, p is fixed. According to Equation (3.1)(3.3)(3.4)(3.6), Rs is almost

unchanged and R P'
P, +Pc
It also can be observed that the maximum throughput is higher in the case that RTS/CTS

is not used than in the case that RTS/CTS is used, no matter how large the payload size is.

This is because that the maximum throughput is obtained when p is relatively small and

thus the impact of collisions due to long data packets could be ignored. As a result, if

RTS/CTS is not used, the MAC overhead is reduced, which results in higher throughput.

On the contrary, in the saturated case where the collision probability is much higher, the

use of RTS/CTS does improve the throughput, especially when the payload size is large.

This is because the impact of collisions due to long data packets becomes significant in












0.8

channel utilizationj- -
0.6
-6-~ - - -____-_- --

0.4

maximum w RTS/CTS
0.2- normalized throughput maximum w/o RTS/CTS
Saturated w RTS/CTS
saturated w/o RTS/CTS
0
0 200 400 600 800 1000 1200 1400 1600
payload size (bytes) (n=100)

Figure 3-3: Impact of payload size and the RTS/CT, mechanism


the saturated case and cannot be ignored; the exchange of RTS/CTS avoids long packet

collisions and thus reduces MAC overhead. Note that for the payload size that is shorter

than about 220 bytes in this parameter setting, the use of RTS/CTS is counterproductive

because of its relatively high overhead compared with the short payload size.

To sum up, to maximize the system throughput, the basic access without the RTS/CTS

mechanism is desired, as long as we can keep the collision probability at a relatively small

value.

3.3.2 Delay and Delay Variation

In this subsection, we study the delay and delay variation performance, which is an in-

tegral part of QoS provisioning in the 802.11 WLAN. As we know, the delay in the network

comprises three components, i.e., propagation delay, transmission delay, and queueing de-

lay. Note that in the WLAN, transmission delay contains a variable amount of delay caused

by MAC layer collisions and thus is not fixed. Henceforth, we define the sum of the propa-

gation delay and transmission delay as the service time at the MAC layer, which is the time

period from the instant that a packet begins to be serviced by the MAC layer to the instant

that it is either successfully transmitted or dropped after several failed retransmissions.









In the following, we will give an analysis of the service time and the queueing delay.

Then, the estimates of delay and delay variation are derived.

Service time distribution

Markov Chain Model for the Service Time

After examining the transmission procedure introduced in section 3.2.1, we can con-

clude that the only outside factor is the collision probability p when the node attempts

the transmission. As discussed in the previous section, p is determined by the number of

neighboring nodes and the traffic distribution at those nodes. Thus we could assume that

p is independent of the backoff state of the node under consideration, although it is still

dependent on the backoff states of other nodes. We therefore can model the stochastic

process of the service time as a Markov chain, since the future state only depends on the

current state. Clearly, the transition probabilities are dependent on the collision probability

p, thus the service time distribution is a function of p.

Probability Generating Function of the Service Time

The service time for each packet consists of multiple backoff time slots which could

be empty slots, collision slots, or successful transmission slots. As mentioned earlier, since

the length of an empty backoff slot is a fixed value and T,,, or Tco depends on the length of

the header and data packet, which are discrete in bits, it is suitable to model the service time

distribution as a discrete probability distribution. To facilitate analysis, this distribution is

completely described with the probability generating function (PGF).

By applying the signal transfer function to the generalized state transfer diagram of

Markov Chain, we have derived the PGF of the service time, GTs(Z), which is quite accu-

rate as verified by ns-2 simulations ([154, 160]). On the other hand,

i00
0pi_ Z 'i, (3.11)

where tsi(i > 0) are all possible discrete values of service time Ts and pi = Pr{Ts

tsi}. We also found that given p, the service time distribution is almost insensitive to









60
payload size = 8000bits, with RTS/CTS
50

E 40 Mean
E 3 Stanard Deviation
30

S20-
U)
10 -
10 G 19 --- e

10-4 10-3 10-2 10-1
Collision probability

Figure 3-4: Mean and standard deviation of service time


n, while n only influences the maximum value of p as shown in Fig. 3-2(a). Thus, the

following delay analysis is valid for different n and we need not specify the value of n.

Mean and Variance of the Service Time

Given Equation 3.11, it is easy to obtain any moment of the service time Ts by taking

the derivative of Grs(Z) with respect to Z. Specifically, the mean and variance are


E[Ts]= LGTs(Z) z= G'(1)(
(3.12)
VAR[Ts] = GC,(1) + G',(1) [G',(1)]2

Fig. 3-4 demonstrates the mean and variance of the service time as a function of the

collision probability p. It can be seen that when p > 0.1, both the mean and the variance

increase exponentially with p. On the other hand, we have found that when p < 0.1,

the achieved throughput is almost the same as the maximum achievable throughput. To

provide a delay guarantee for some delay-sensitive applications such as voice over IP, and

achieve approximately maximum throughput, the wireless LAN should keep the collision

probability less than 0.1.

Packet delay bound and delay variation estimate

Because there is typically one shared outgoing queue for all packets from different

applications in each mobile node, we can model each node as a queueing system. In the

queueing system, the packet arrival process is determined by the aggregate traffic behavior









of all applications that emit packets to the MAC layer; the service time follows the dis-

tribution described in previous subsection. After building such a queueing model, we can

derive accurate estimates of delay and delay variation in the non-saturated case. Notice that

the number of packets waiting in the queue, Nq, almost equals zero in the non-saturated

case especially forp < 0.1 as shown in the papers [154, 160] and verified in our simulation

later. Otherwise each node will contend for the channel in most times and result in a much

higher p.

Delay Bound with Known Packet Arrival Rate

We start the analysis with a simple case, i.e., the packet arrival follows some process

with a known (or estimated) arrival rate. If the arrival process is Poissonian, the system can

be modeled as a M/G/1 system [86]. Accordingly, the mean of the packet delay T, which

consists of the waiting time in the queue and the service time, is

ATs2
E[T] = Ts + (3.13)
2(1 p)

where A is the average arrival rate of the input traffic and p = A x Ts < 1. If the arrival

process follows a general distribution, then we get a G/G/1 system, for which we have an

upper bound Tu for T [87],


E[T] < Ts + T, T (3.14)
2(1 p)

where T s and a, are the standard deviations of the service time and packet arrival process,

respectively.

So far, these results hold when p < 1 for the system with infinite buffer. The actual

delay upper bound should be less than Tu because we do not count the packets dropped

due to limited buffer, which will have a long delay in the system with infinite buffer. In

fact, because we are only interested in the non-saturated case with an almost empty queue,

the above results are thus accurate.

Delay Bound and Delay Variation with Unknown Packet Arrival Rate









In the previous paragraph we only give the mean of the packet delay T with the es-

timation dependent on the specific packet arrival process and on the accurate estimate of

A. In reality, however, this approach could be infeasible if A is hard to estimate when the

instantaneous packet arrival rate at each individual node changes dramatically. We thus

embark on deriving the accurate estimates for delay and delay variation in a more general

case, i.e., without any knowledge about A.

Let Tsi denote the service time for the i-th packet at a node under consideration.

Since the backoff timer is reset for every packet to be transmitted [68], {Tsi, Ts2,...} are

iid (independently and identically distributed) random variables. Let Ti be the system time

(or delay) of the i-th packet including the service time and the waiting time in the queue,

Ri be the residual service time seen by the i-th packet, and Ni be the number of packets

found waiting in the queue by the i-th packet at the instant of arrival.

Based on such notations, we obtain
i-1
T, =Tsi + Ri + Tsj (3.15)
j i-Ni

As previously discussed, Ni almost equals 0 in the non-saturated case, so we can

approximate Ti as

Ti Tsi + Ri (3.16)

Notice that Ri is the residual service time of the (i Ni 1)-th packet, thus we have


Tsi Ti Tsi + TSiN,- (3.17)


By taking expectations on both sides of Equation (3.17) we have


E[Ts] < E[T] < 2E[Ts]


(3.18)








Since it is difficult to derive the variation of Ri in general cases, we use the standard devi-
ation of the service time o to approximate that of T, i.e., oT as follows:

Ts
where k is a constant value. From ns-2 simulation results as presented later, k = 1, or 2
gives a good approximation.
In fact, by applying the Residual Life Theorem [86], we could obtain more accurate
approximations of E[T] and Ty. Let r be the residual service time observed at any time
instant during the service. If the service time distribution is FTs(x), then the pdf of r,
denoted by f,(x), can be expressed as p(l FTs(x)), where p = We thus have

E[r] =fo rfr(x)dx = E[Ts2]
E[r2] 0o r2(x)dx = E[Ts3]
E[R] 0 x P(idle) + E[r]P(',, ,,i) = E[Ts2] (3.20)
E[R2] 0 x P(idle) + E[r2]P(1,, ,,) = jE[Ts3]
Var[R]= -"E[Ts3] ("E[Ts2])2

where ro = P(',, ,ii) is the probability that the server is busy, i.e., there is one packet
contending for the channel or being transmitted. Because ro < 1, we obtain

E[Ts2]
E[T] E E[Ts] + E[R] < E[Ts] + TUR (3.21)
2E[Ts]

Var[R] = (-T) [Ts3] + r2 (E[Ts3] (E[TS2])2)
(-o E[Ts3] + rVar[r] (3.22)

4 E[Ts3] + Var[r] = E[Ts31 (LE[Ts2])2

Var[T] r Var[Ts] + Var[R] (3.23)
(3.23)
<,Var\Ts] + 5E[TS3 -_ (E 2 g2
12E[Ts] 2E[Ts] TUR









Fig. 3-5(a) illustrates both the lower bound and the upper bound for the packet delay

T. We can see that the upper bound and lower bound are very close, thus we can charac-

terize the delay with high accuracy, although the exact value is not available. As expected,

when p < 0.1, TUR is tighter than 2 x E[Ts]. This is desirable since we focus on the

non-saturated case where p is small. As revealed by the bounds, the mean of the system

delay T is small: 5ms < E[T] < 10ms when p < 0.01, and E[T] < 30ms when p < 0.1.

This is sufficient for real-time applications such as VolP.

The standard deviation of the system delay is illustrated in Fig. 3-5(b). As shown

in the figure, it is also small: aJT < 30ms when p < 0.1. When p < 0.02, the standard

deviation is much smaller than E[Ts] (and than E[T] since E[Ts] is the lower bound).

Note that TUR,, is relatively large when p < 0.002. This is because the approximation

in Equation (3.22) uses ro > 0.5; however, ro should have been smaller than 0.5 when

p < 0.002.

As a special case, if the packet arrival process is Poissonian, then ro = p = AE[Ts] <

1. Thus

E[T] E E[Ts] + E[R] = E[Ts] + -AE[Ts2] TURM, (3.24)
2

Var[T] w Var[Ts] + Var[R]
(3.25)
Var[Ts] + AE[Ts3] (AE[Ts21)2 a-2

Finally, we comment on the results of delay and delay variation. First, all the above

results are derived for the non-saturated case, which means the traffic intensity p < 1 and

the collision probability p < 0.1. Second, the approximation in Equation (3.16) relies on

the assumption that there is no bulk arrival. Although this assumption is common in the

analysis of queueing systems and is true for both the Poisson arrival process and determin-

istic arrival process, in practice, bursty traffic such as TCP traffic violates it. Consequently,

the bursty traffic induces not only longer waiting time in the queue, but also higher colli-

sion probability in the burst period leading to longer service time. For the above results to

remain valid, it is necessary to regulate arriving traffic at the MAC layer.











payload size = 8000 bits, with RTS/CTS
10o
S- E[Ts]
- 2 x E[Ts]
S101 TUR


10-2
-u 10 --: -- -
Q l _ .. --


10-4 10-3 10-2
collision probability p
(a) Delay bound


10-1 100


GTs
2 x (Ts
3 x Ts
(T
UR


payload size = 8000 bits, with RTS/CTS

10 103 102 10-1
collision probability p
(b) -": .. i i deviation of delay

U.- .:: .. 3-5: Packet delay


3.3.3 Packet Loss Rate

At the MAC layer, a packet may be lost due to queue overflow or MAC collisions.

Once a packet is queued, a node attempts to transmit it for a certain number of times,

denoted by a. If all the attempts fails due to collisions, the packet gets dropped. Given the

collision probability p, the packet dropping probability Pd due to MAC collisions is


Pd = p,


(3.26)


When the packet blocking probability Pblock, i.e., the probability that the queue is

full when a packet arrives, is very small, as for the non-saturated case where Nq 0 as

mentioned earlier, the total packet loss rate PI of the queueing system can be approximated


10-4
10-5









as Pd, i.e.,

Pl Pa (3.27)

We see when p < 0.1 and a = 7 [68], Pl < 10-7. Obviously, this satisfies the packet

loss requirements of most applications such as VoIP.

On the contrary, a much higher packet loss rate is expected if the network is in the

saturated case for the following reason. On one hand, the collision probability p gets sig-

nificantly large, resulting in considerable packet losses due to collisions. On the other hand,

each packet experiences a much longer system delay in the saturated case compared to that

in the non-saturated case, which leads to a full queue at most times and hence blocks newly

arriving packets.



Before ending this section, we make a few remarks about the analytical model. Note

that all the performance metrics are expressed as a function of the collision probability.

However, obtaining the collision probability is not easy. There are two possible approaches.

One is to analytically derive the collision probability, which requires the full knowledge of

the traffic arrival models at the node of interest and at all the other nodes in the network

as well. The other is to measure it through experiments. Unfortunately, it is not amenable

to practical measurement due to the lack of measured values or the inability of each node

to distinguish collisions from channel fading. Therefore, we propose the channel busyness

ratio as a good substitute for the collision probability for the following reasons. First,

as mentioned earlier, the channel busyness ratio is an injective function of the collision

probability. This indicates that the channel busyness ratio can also serve as the input of

the analytical model. Unlike the collision probability, the channel busyness ratio is easy

to measure in practice because the IEEE 802.11 is essentially based on carrier sensing.

Second, as shown for the non-saturated case, the channel busyness ratio can accurately

represent the channel utilization or the normalized throughput, and hence can be used to

facilitate network control mechanisms such as call admission control over the real-time









traffic and rate control over the best effort traffic. Accordingly, all the performance metrics

are presented as a function of the channel busyness ratio in the following simulation results.

3.4 Simulation Study of the IEEE 802.11

The simulation study in this section serves two purposes. First, it is aimed at verifying

our analytical study in section 3.3. Second, while our analytical results have shown that the

IEEE 802.11 can operate at an optimal point that leads to maximum throughput, low delay,

and almost zero packet loss rate, they do not reveal a specific way to achieve this optimal

operating point. Thus we demonstrate how to reach and retain the optimal point through

simulations.

3.4.1 Simulation Configuration

The simulation study is conducted using the ns-2 simulator. The IEEE 802.11 system

parameters are summarized in Table 3-2. The RTS/CTS mechanism is used. We simulate

different number of mobile stations in the wireless LAN. Every node initiates an identical

UDP/CBR traffic flow to a randomly selected neighbor. The queue length at each node is

10 packets.

As revealed, whether the network operates in the non-saturated or saturated case can

be determined by controlling the collision probability p. Also, the optimal operating point

lies where p t 0.1. Without changing the 802.11 protocol, we use two techniques to

control p in order to locate the optimal point. One is to schedule the start time of the UDP

flows, which will be described below; the other is to gradually increase the sending rate of

each flow from 0. In contrast, the saturated case can be easily simulated by boosting the

traffic load to a much higher level than what the network can support.

Deterministic minimum-collision-probability scheduling (DPS)

To minimize the collision probability, we schedule UDP flows in such a way that

start time of one flow is separated from another by a constant period tint/n, where tint is

the packet inter-arrival time for each flow. So if the aggregate traffic rate is less than the

network capacity, i.e., the network can handle all the arriving packets from each flow, the













= -+- DPS
E 0.6 Saturated Case
N Analysis
00.5 -
E
z 0 50 100 150 200 250 300
(a) number of nodes n

E DRS
10 DPS


0 50 100 150 200 250 300
(b) number of nodes n
20 r
10o1 ---------"r------ I
S------ Saturated Case

0 50 100 150 200 250 300
(c) number of nodes n

Figure 3-6: Simulation results when payload size = -.-i.i


collision probability could be reduced to zero. In this case, there is no queueing delay and

the system delay is the random backoff time plus one packet transmission time. We call

this scheduling deterministic minimum-collision-probability scheduling.

Distributed randomized scheduling (DRS)

However, in a distributed WLAN environment, it is very difficult for each node to

exactly know the start time of all the flows and schedule its own flows accordingly to

avoid collisions. Therefore, to simulate a more realistic scenario, we cannot adopt the de-

terministic scheduling described above. We thus employ a simple yet effective scheduling

algorithm that starts each flow at randomized times. Specifically, the start time of each flow

is uniformly chosen in [0, tit], which keeps all the nodes from contending for the channel

at the same time. As a result, the collision probability is reduced and no node needs to care

about other nodes' transmission schedule.










08 10
Simulation E[T] 100 T
0 Analysis -- E[Ts] T
S 06 2 x E[Ts] __ x T
06 10 T-UR
/ 5T 102

03 / '
010
01 10 10
0 103
0 02 04 06 08 1 0 02 04 06 08 1 0 02 04 06 08
(a) Channel busyness ratio (b) Channel busyness ratio (c) Channel busyness ratio

Figure 3-7: Simulation results when n=50 and payload size = 8000bits


3.4.2 Simulation Results

Saturated case vs. non-saturated case

In Fig. 3-6, for the non-saturated case, we see that the normalized throughput that

DPS achieves is slightly higher than the theoretical maximum throughput, since it uses

perfect scheduling and hence reduces the collision probability to zero. Likewise, the nor-

malized throughput that DRS achieves is close to the theoretical maximum throughput,

since it greatly reduces the collision probability. On the contrary, the throughput in the

saturated case is much lower. As is consistent with the analytical results, the non-saturated

throughput is almost independent of the number of nodes, whereas the saturated through-

put declines significantly with the increase in node number. For delay, we see that there is

difference in orders of magnitude for these two cases. Also, while the delay stays almost

unchanged in the non-saturated case as the number of node increases, it increases in the

saturated case. This is due to the fact that in the latter case, each node always has packets to

transmit and keeps contending for the channel, which greatly increases the collision prob-

ability. As a result, each packet suffers from both long queueing delay and service time.

Note that DPS enjoys a shorter delay than DRS since it reduces the collision probability

more effectively.

Optimal operating point

As Fig. 3-6 shows, DRS yields a comparable performance with that of DPS, we thus

use DRS as our scheduling algorithm henceforth. By gradually increasing the sending rate









of each flow, we are able to locate the optimal operating point as shown in Fig. 3-7 and

3-8. While Fig. 3-7 presents the performance of throughput, delay, and delay variation

as a function of the channel busyness ratio, Fig. 3-8 shows the behavior of average queue

length and packet loss rate when input traffic increases.

Two important observations are made. First, we observe there is a turning point in

all the curves where the channel busyness ratio is about 0.95. Before that point, as the

input traffic increases, the throughput keeps increasing, the delay and delay variation are

small and almost unchanged, the queue at each node is empty, and the packet loss rate is

zero. Note that with the small delay and delay variation, the delay requirements of the

real-time traffic can be adequately supported. After that point, the queue and the collision

probability forms a positive feedback loop. A slightly larger collision probability causes

the queue to build up. The queue, even with one packet always in it, will force the MAC to

keep contending for the channel, thereby exponentially increasing the collision probability,

which in turn forces more packets to accumulate in the queue. Then, catastrophic effects

take place: the throughput drops quickly, the queue starts to build up and the delay and

delay variation increase dramatically, and the packet suffers from a large loss rate. Clearly,

this turning point is the optimal operating point that we should tune the network to work

around, where the throughput is maximized and the delay and delay variation are small.

Second, as shown in Fig. 3-7, the simulation results verify our analytical study of the

IEEE 802.11. The throughput curves obtained from analysis and simulation coincide with

each other. Also as indicated in our analytical study, before the optimal point is reached,

the network stays in the non-saturated case and the queueing delay is almost zero; thus

the packet delay T can be accurately estimated by the service time Ts, which provides

the lower bound. Meanwhile, the mean and variation are well bounded by TUR and oTR

before the turning point as shown in equations (3.18, 3.19, 3.21, 3.23).










F 100
0)


S50
o-
0)

0 1
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
(a) Channel busyness ratio


t)
0 0.5-

0
n 0

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
(b) Channel busyness ratio

Figure 3-8: Simulation results when n=50 and payload size = 8000bits


3.5 Discussions

3.5.1 Impact of Fading Channel

So far it is assumed the channel is perfect. However, when channel fading is figured

in, packet losses are no longer due to collisions only; they may well be caused by channel

fading. Practically, it is extremely difficult to distinguish these two causes. As a matter

of fact, the 802.11 responds in the same way if the transmitter does not correctly receive

its expected frame, which may be either CTS or ACK, no matter whether this is due to

collision or channel fading. Based on this observation, we can incorporate the packet error

probability into the collision probability as the recent work [58] did, and all the analytical

results still hold.

It is important to note that normally channel fading is not a serious problem in the

WLAN, which features low node mobility and relatively stable channel. However, if the

packet error probability due to channel fading becomes significant, i.e., the equivalent col-

lision probability is high in our model, the QoS level will be hurt. Our analytical results

show that in this case, as illustrated in Fig. 3-1, 3-4, 3-5(a), and 3-5(b), the normalized

throughput decreases, the service time increases, the mean and variation of delay increase









along with the service time, and packet loss rate increases as well. However, with our an-

alytical model, we can still calculate the maximum throughput, packet loss rate, and give

accurate estimates of delay and delay variation according to Equations (3.5), (3.18), (3.19),

(3.21), and (3.23).

3.5.2 Impact of Prioritized MAC

Since our focus is on how well the original IEEE 802.11 DCF can support QoS, we do

not change the MAC protocol in the analysis and simulations. Within either the real-time

traffic or the best effort traffic, no differentiation is committed. As a result, all the real-time

traffic, including CBR and VBR traffic, equally shares the delay and delay variation, which

sometimes is not flexible enough. If a prioritized 802.11 MAC protocol similar to [1, 125]

is adopted, we are able to provide priority within the real-time traffic. As a result, the high

priority real-time traffic receives smaller delay variation, whereas the low priority real-time

traffic receives higher delay variation [33].

3.6 Conclusion

Despite considerable efforts spent on performance analysis and QoS provisioning for

the IEEE 802.11 WLAN, the question of how well it can support QoS remains vague. In

this chapter, we clearly answer this question through thorough studies, which constitutes

our key contribution.

We have analytically characterized the optimal operating point for the 802.11 WLAN,

and shown that if the network is tuned to work at this point, in addition to achieving theoret-

ical maximum throughput, it can support the major QoS metrics such as throughput, delay

and delay variation, and packet loss rate, as required by real-time services. This is further

validated via extensive simulations. We therefore clarify that the IEEE 802.11 WLAN can

provide statistical QoS guarantees, not just differentiated service, for multimedia services.

We also demonstrate that the channel busyness ratio can accurately and timely represent

the network utilization; hence it can be used to facilitate the regulation of total input traffic

to support QoS.















CHAPTER 4
A CALL ADMISSION AND RATE CONTROL SCHEME FOR MULTIMEDIA
SUPPORT OVER IEEE 802.11 WIRELESS LANS

Quality of service (QoS) support for multimedia services in the IEEE 802.11 wireless

LAN is an important issue for such WLANs to become a viable wireless access to the

Internet. In this chapter, we endeavor to propose a practical scheme to achieve this goal

without changing the channel access mechanism. To this end, a novel call admission and

rate control (CARC) scheme is proposed. The key idea of this scheme is to regulate the

arriving traffic of the WLAN such that the network can work at an optimal point. We first

show that the channel busyness ratio is a good indicator of the network status in the sense

that it is easy to obtain and can accurately and timely represent channel utilization. Then we

propose two algorithms based on the channel busyness ratio. The call admission control

algorithm is used to regulate the admission of real-time or streaming traffic and the rate

control algorithm to control the transmission rate of best effort traffic. As a result, the real-

time or streaming traffic is supported with statistical QoS guarantees and the best effort

traffic can fully utilize the residual channel capacity left by the real-time and streaming

traffic. In addition, the rate control algorithm itself provides a solution that could be used

above the media access mechanism to approach the maximal theoretical channel utilization.

A comprehensive simulation study in ns-2 has verified the performance of our proposed

CARC scheme, showing that the original 802.11 DCF protocol can statically support strict

QoS requirements, such as those required by voice over IP or streaming video, and at the

same time, achieve a high channel utilization.

4.1 Introduction

In recent years, the IEEE 802.11 wireless LAN [68] has been increasingly employed

to access the Internet because of its simple deployment and low cost. According to the









IEEE 802.11 standard, the medium access control (MAC) mechanism contains two access

methods, i.e., Distributed Coordination Function (DCF) and Point Coordination Function

(PCF), with the former being specified as the fundamental access method. Despite its

popular use, currently only best effort traffic is supported in DCF. Section 4.2 describes the

802.11 protocol in more detail.

Quality of service (QoS) provisioning for multimedia services including voice, video,

and data is crucial for the IEEE 802.11 wireless LAN to continue to thrive and evolve as a

viable wireless access to the Internet. Although there are several schemes ([96, 9, 88, 34,

124]) which use PCF mode to support QoS for real-time traffic, we do not discuss further

along this line because PCF is an optional access method ([68]) which is only usable on

infrastructure network configurations and not supported in most current wireless cards. In

addition, it may result in poor performance as shown in the papers [94, 145, 126]. In

this chapter, we focus on the 802.11 DCF mode. However, guaranteeing QoS for real-time

traffic in the 802.11 DCF mode is not an easy task given that it is in nature contention-based

and distributed, and hence render effective and efficient control very difficult. Furthermore,

other problems such as hidden terminals or channel fading make things worse.

In face of these challenges, considerable research ([1, 81, 107, 114, 119, 125, 137]) has

been conducted to enhance the IEEE 802.11 WLAN to support service differentiation or

prioritized service [18]. Ada and Castelluccia [1] proposed to scale the contention window,

use different inter frame spacing or maximum frame length for services of different priority.

As a matter of fact, similar ideas have recently been adopted in the enhanced DCF (EDCF)

defined in the IEEE 802.1 le draft ([72, 31, 99]). two mechanisms [125], i.e., virtual MAC

and virtual source, were proposed to enable each node to provide differentiated services for

voice, video, and data. By modifying the 802.11 MAC, a distributed priority scheduling

scheme was designed to approximate an idealized schedule, which supports prioritized

services [81]. Similarly, by splitting the transmission period into a real-time one and a

non-real-time one, real-time traffic is supported with QoS guarantee [114]. However, the









DCF mode was dramatically changed. The Blackbust [119] provided high priority for real-

time traffic. Unfortunately, it imposes special requirements on high priority traffic and is

not fully compatible with the existing 802.11 standard. In summary, if the semantics of the

802.11 DCF is maintained, only differentiated service, rather than stringent QoS assurance,

is supported.

Meanwhile, much effort has also been spent in improving throughput for the 802.11

DCF ([12, 13, 16, 20, 85, 90]). Based on the work [21], Cali et al. attempted to approach

the protocol capacity by replacing the exponential backoff mechanism with an adaptive one

[20]. Kim and Hou developed a model-based frame scheduling algorithm to improve the

protocol capacity of the 802.11 [85]. Two fast collision resolution schemes were proposed

by Bharghavan [13] and Kwon et al. [90], respectively. The idea is to use two channels or to

adjust backoff algorithms to avoid collisions, thereby providing higher channel utilization.

It is important to note that all these works focused on the throughput in the saturated case.

In our previous work [150], We have shown through both theoretical and simulation

studies that the IEEE 802.11 DCF protocol could satisfy the QoS requirements of the real-

time and streaming traffic while achieving the maximal channel utilization when it is work-

ing at the optimal point corresponding to a certain amount of arriving traffic. If the arriving

traffic is heavier than this threshold, the WLAN enters saturation, resulting in significant

increase in delay and decrease in throughput; on the other hand, if the arriving traffic is less

than this threshold, channel capacity is wasted. In reality, however, to tune the network that

operates on the basis of channel contention to work at this point requires an effective and

efficient control algorithm to regulate the input traffic [109]. Therefore, we are motivated

to design a call admission and rate control scheme (CARC) (Section 4.4). Specifically, call

admission control (CAC) is used for real-time or streaming traffic, and rate control (RC)

for best effort data traffic.

Essentially, the CARC scheme has the following distinguishing features:









It utilizes an new measure of network status, the channel busyness ratio to exercise

traffic regulation, which is easy to obtain and can accurately and timely represent the

network utilization as shown in Section 4.3.

The call admission control scheme is able to provide statistical QoS guarantees for

real-time and streaming traffic.

The rate control scheme allows best effort traffic to utilize all the residual channel

capacity left by the real-time and streaming traffic while not violating their QoS

metrics, thereby enabling the network to approach the maximal theoretical channel

utilization.

Since each node keeps track of the channel busyness ratio locally to conduct control,

this scheme is distributed and suits well with the DCF mode.

We have implemented the CARC scheme in ns-2 [106], and conducted a comprehen-

sive simulation study to evaluate its performance. As shown in Section 4.5, CARC is able

to support real-time services, such as voice and video, with QoS guarantees, and achieve

high throughput by allowing best effort traffic to make full use of the residual channel ca-

pacity. This confirms that the 802.11 WLAN cannot only support differentiated service,

but also support strict QoS.

In Section 4.6, we discuss the effect of channel fading on our scheme and the possible

implications arising due to the employment of a prioritized 802.11 DCF. Finally, Section

4.7 concludes this chapter.

4.2 Background

4.2.1 Operations of the IEEE 802.11 DCF Protocol

The basic access method in the IEEE 802.11 MAC protocol is DCF (Distributed coor-

dination function), which is based on carrier sense multiple access with collision avoidance

(CSMA/CA). Before starting a transmission, each node performs a backoff procedure, with

the backoff timer uniformly chosen from [0, CW] in terms of time slots, where CW is the

current contention window. If the channel is determined to be idle for a backoff slot, the









backoff timer is decreased by one. Otherwise, it is suspended. When the backoff timer

reaches zero, the node transmits a DATA packet. If the receiver successfully receives the

packet, it acknowledges the packet by sending an acknowledgment (ACK) after an inter-

val called short inter-frame space (SIFS). So this is a two-way DATA/ACK handshake. If

no acknowledgment is received within a specified period, the packet is considered lost; so

the transmitter will double the size of CW and choose a new backoff timer, and start the

above process again. When the transmission of a packet fails for a maximum number of

times, the packet is dropped. To reduce collisions caused by hidden terminals [14], the

RTS/CTS (request to send/clear to send) mechanism is employed. Therefore, a four-way

RTS/CTS/DATA/ACK handshake is used for a packet transmission.

In the IEEE 802.11, the network can be configured into two modes, i.e., infrastructure

mode or ad hoc mode. In the infrastructure mode, an access point (AP) is needed to partic-

ipate in the communication between any two nodes, whereas in the ad hoc mode, all nodes

can directly communicate with each other without the participation of an AP.

4.2.2 QoS Requirements for Multimedia Services

As the Internet expands its supported traffic from best effort data to a variety of mul-

timedia services, including video conferencing, voice over IP (VoIP), streaming audio and

video, WWW, e-mail, and file transfer, etc., QoS provisioning has become an important

issue. The commonly accepted QoS metrics mainly include bandwidth, delay, delay jitter

(i.e., dalai variation), packet loss rate (or bit error rate). According to their QoS require-

ments, current multimedia services can be grouped into three classes: real-time, streaming,

and non-real-time (or best effort).

Real-time: Real-time traffic has stringent requirements in delay and delay jitter,

which is necessary for interactive communications like VoIP and videoconferencing. Ac-

cording to the ITU standards [73, 74], the one way transmission delay should be prefer-

ably less than 150ms, and must be less than 400ms. However, it is not very sensitive to

packet loss rate. Typically, a loss rate of 1% is acceptable for real-time video with rate









16 ~ 384Kbps and a loss rate of 3% for real-time audio with rate 4 ~ 64Kbps. Because

delayed packets are not tolerable, retransmission of lost packets is not useful. Thus, UDP

is used to transmit real-time traffic.

Streaming: Streaming audio or video belongs to this class. Compared with real-time

traffic, it is less sensitive to delay or delay jitter. At the expense of increased delay, playout

buffer (or jitter buffer) can be used to compensate for delay jitter in the range of 20 ~ 50

ms. As specified in the ITU standard ITU-G1010 [74], acceptable delay may be up to 10

seconds, while the packet loss rate is about 1%. Streaming traffic is normally transported

via UDP, although a retransmission strategy can be added in the application layer.

Non-real-time: Non-real-time services comprise e-mail, file transfer, and web brows-

ing. Most non-real-time services are tolerant to delay ranging from seconds to minutes or

even hours. However, the data to be transferred has to be received error-free, which means

reliable transmission is required. So non-real-time traffic is transported with TCP.

4.3 Channel Busyness Ratio

In this section, we give the definition of the channel busyness ratio and elaborate on

why and how it can be used to represent the network status.

4.3.1 Definition of Channel Busyness Ratio

At the MAC layer, a backoff time slot could be an empty slot, a period associated with

a successful transmission, or a period associated with a collision ([68, 15, 154, 160]). Let

pi, ps, and pc be the probabilities that the observed backoff time slot is one of the three kinds

of slots, respectively. Let Tsu, be the average time period associated with one successful

transmission, and Too be the average time period associated with collisions. Then

Tsu, = rts + cts + data + ack + 3sifs + difs
(4.1)
Tool = rts + cts_timeout + difs = rts + eifs









for the case where the RTS/CTS mechanism is used, and

T,, = data + ack + sifs + difs (4.2)
(4.2)
Tcol = data* + acktimeout + difs = data* + eifs

for the case where there is no RTS/CTS mechanism, where data and data* (please refer

to [15] for derivation of data*) are the average length, in seconds, for the successful trans-

mission and collision of the data packets, respectively. Notice that the sources keep silent

when waiting CTS packets, and any station which senses a collision will set the network

allocation vector (NAV) [68] with an eifs period. Thus, it can be easily obtained that

Ri pio
Pi,++PsTsuc+PcTcol

Rb 1 R, (4.3)

s Pi~~sTs Uc+cTco

where a is the length of an empty backoff time slot, Ri is defined as the channel idleness

ratio, Rb the channel busyness ratio, and R, the channel utilization. Clearly, the channel

busyness ratio Rb can also be thought of as the ratio of time that the channel is busy due

to successful transmissions as well as collisions to the total time. Once we obtain Rs, the

normalized throughput s is expressed as


s = Rs x data/T,,suc, (4.4)


and the absolute throughput is s times the bit rate for data packets.

4.3.2 Channel busyness ratio: an accurate sign of the network utilization

First, we build the relationship between the channel busyness ratio and the packet

collision probability, denoted by p, that a node may experience.









We assume the total number of nodes in a WLAN is n. The transmission probability

for each node in any backoff time slot is pt. Obviously, we obtain the following equations:

pI = (1 pt)n

ps = npt(l pt)n (4.5)

P = 1 i Ps

Meanwhile, p can be expressed in terms ofpt as follows:


p =1 (1 pt)n-1 (4.6)


According to Equation (4.3)(4.5)(4.6), we can express Rb, Rs, and s as a function of

p, which are shown in Fig. 4-1. All the parameters involved are indicated in Table 4-1 and

most are the default values in the IEEE 802.11. In Fig. 4-1, three cases, i.e., n = 5, 10,

and 300, are considered.

Several important observations are made for Fig. 4-1. First, we find that the channel

busyness ratio is an injective function of the collision probability. In fact, this can easily be

proved. Second, when p < 0.1, Rb is almost the same as Rs, namely


R, 5 Rb. (4.7)

This is not hard to understand. When the collision probability p is very small, the channel

resource wasted in collisions is so minor that it can be ignored. Third, the maximal through-

put is almost insensitive to the number of active nodes. As a matter of fact, we have shown

in our previous work [150] that the point where the maximal throughput is achieved is the

optimal working point for the network where the collision probability is very small and the

packet delay and delay jitter are small enough to support the QoS requirements of real-time

traffic. Given these observations and the fact that the throughput is proportional to Rs as

shown in Equation (4.4), we therefore could use the measured channel busyness ratio Rb

to accurately estimate the throughput from zero to the maximum value.









RTS/CTS scheme with different number of nodes

n=5 .
0.8 n=10 channel busyness ratio\
n=300
0.6 channel utilization

0// normalized t roughput
0.4 -


0.2 -


0
104 10-3 10-2 10-1 100
collision probability p

Figure 4-1: Channel busyness ratio and utilization


Next, we present some ns-2 simulation results in Fig. 4-2, which shows the perfor-

mance of throughput, delay, and delay variation as a function of the channel busyness ratio.

Again, the IEEE 802.11 system parameters are summarized in Table 4-1. Every node initi-

ates an identical UDP/CBR traffic flow to a randomly selected neighbor. The queue length

at each node is 100 packets. Different points in Fig. 4-2 corresponds to different sending

rate of flows. It can be seen that there is a turning point in all the curves, where the channel

busyness ratio is about 0.95. Before that point, as the input traffic increases, the throughput

keeps increasing, the delay (including queueing delay, backoff time and transmission time)

and delay variation does not change much and is small enough to support the real-time

traffic. After that point, the throughput drops quickly and the delay and delay variation

increase dramatically. Clearly, this turning point is the optimal operating point that we

should tune the network to work around, where the throughput is maximized and the delay

and delay variation are small. Therefore, the network status is known by keeping track of

the channel busyness ratio.

Further, if we denote by Bu the channel utilization corresponding to the optimal point,

we can estimate the available normalized throughput by s, = (Bu Rb) x data/T",u before

the network achieves the maximal throughput. As shown in our work [150], Bu is almost










Table 4-1: IEEE


71

802.11 system parameters


Bit rate for DATA packets 2 Mbps
Bit rate for RTS/CTS/ACK 1 Mbps
PLCP Data rate 1 Mbps
Backoff Slot Time 20 ps
SIFS 10 ps
DIFS 50 ps
Phy header 192 bits
MAC header 224 bits
DATA packet 8000 bits + Phy header + MAC header
RTS 160 bits + Phy header
CTS, ACK 112 bits + Phy header


S0.6

- 0.4
.N
E
S0.2
Z


(a) Channel busyness ratio


10
0 0.2 0.4 0.6 0.8 1
(b) Channel busyness ratio


0 0.2 0.4 0.6 0.8 1
(c) Channel busyness ratio


Figure 4-2: Simulation results when number of nodes equals 50 and RTS/CTS mechanism
is used


the same for different number of active nodes and packet size, and Bu a 0.90 (without

RTS/CTS) or Bu w 0.95 (with RTS/CTS).
4.3.3 Measurement of Channel Busyness Ratio

According to the definition of Rb, it is easy to conduct the measurement since the

IEEE 802.11 is a CSMA-based MAC protocol, working on the physical and virtual carrier

sensing mechanisms. The channel is determined to be busy when the measuring node is

sending, receiving, or its network allocation vector (NAV) [68] indicates the channel is

busy, and to be idle otherwise.
4.4 CARC: Call Admission and Rate Control

As revealed in previous sections, keeping the channel busyness ratio close to a certain

threshold is essential to maximizing network throughput and providing QoS. To accomplish

this goal, it is crucial to regulate total input traffic through call admission control (CAC)









over real-time traffic and rate control (RC) over best effort traffic, given that the 802.11

DCF protocol is designed to provide best effort services and does not differentiate any

types of services.

We thus propose a call admission and rate control (CARC) scheme, which consists

of two mechanisms: CAC and RC. In what follows, the design rationale is discussed first,

followed by detailed descriptions of the CAC and RC algorithm in order.
4.4.1 Design Rationale

The goal of an effective call admission and rate control scheme is to provide QoS for

real-time traffic, and to allow best effort traffic to make full use of the residual channel

resource. In the context of the WLAN where each node only has a partial view of the net-

work, however, the design of CARC is much more complicated than it appears, especially

due to the following difficulties.

The first problem is that multiple new real-time flows may be simultaneously admit-

ted by individual nodes if not coordinated, henceforth referred to as over-admission. To

mitigate this problem, each node can randomly back off to delay a new flow that could be

admitted. During the backoff period, each node keeps monitoring the channel busyness

ratio; if the measured channel busyness ratio is increased (due to the admission of new

flows by other nodes) such that the previously could-be-admitted but delayed new flow can

no longer be accepted, the flow is rejected. Another way is that each node, after admitting

a new flow, drops the flow if later on the measured channel busyness ratio is found to be

greater than the maximum channel utilization. In this case, however, the QoS level of the

real-time flows admitted earlier have already been suffered.

Another more severe issue is that it is very hard for each individual node to accurately

estimate the total traffic rate of the currently admitted real-time flows based on the mea-

sured channel busyness ratio, since the latter also includes the contribution from best effort

traffic. Without an accurate estimate, the rate of best effort traffic cannot be effectively









controlled. This in turn may completely cause the CAC algorithm to reject any real-time

traffic if the channel busyness ratio is boosted to a high level by heavy best effort traffic.

Therefore, to achieve its goal, the CARC scheme must properly address these prob-

lems. To completely avoid the over-admission problem, we opt for a coordinator-aided

CAC scheme. In other words, all admission decisions are made by a coordinating node,

which can record the current number of admitted real-time flows and their occupied chan-

nel bandwidth in the network. Clearly, in this way no over-admission will occur. It is im-

portant to note that a coordinator is available whether the wireless LAN is working in the

infrastructure mode or in the ad hoc mode. If the network is working in the infrastructure

mode, the access point is the coordinator. Otherwise, a mobile node can be elected to act as

the coordinator in the network using one of many algorithms in the literature ([49, 116]).

Further discussions on the election algorithm is beyond the scope of this chapter.

Since the 802.11 DCF is not prioritized, our CAC algorithm guarantees a uniform

QoS level in terms of delay, delay variation, and packet loss rate for all real-time traffic.

Note that two criteria are applied to CAC. The first criterion is that CAC admits a new real-

time flow only if the requested resource is available. Here we need to set an upper bound,

denoted by BM, for bandwidth reservation for real-time traffic [33]. We set BM to 80% (it

could be changed depending on traffic composition) of the maximum channel utilization,

denoted by Bu, of the WLAN for two reasons. It first ensures that the best effort traffic is

operational all the time, since the best effort traffic is at least entitled to 20% of the channel

throughput. In addition, the 20% of the channel throughput for best effort traffic can be

used to accommodate sizable fluctuations caused by VBR real-time traffic. The second

criterion is that the QoS provided for the currently existing real-time flows is not affected.

This can be guaranteed as long as the first criterion is in place to make sure the collision

probability is kept around a small value as shown earlier.

For best effort traffic, the rate control (RC) scheme must ensure two things. First, best

effort traffic should not affect the QoS level of the admitted real-time traffic. Second, best









effort traffic should have access to the residual bandwidth left by real-time traffic in order

to efficiently utilize the channel. Clearly, both demand an accurate estimate of the instan-

taneous rate of ongoing real-time traffic. If the network is working in the infrastructure

mode, this is achievable. In this case, since all communications must go through the access

point, it can monitor the traffic in both directions, i.e., the upstream flows that are from

mobile nodes to the access point, and the downstream flows that are from the access point

to mobile nodes. On the other hand, if the network is working in the ad hoc mode, accurate

rate control becomes much more difficult. In this case, since all mobile nodes can directly

communicate with each other, no node has perfect knowledge of the instantaneous traffic

rate of the real-time traffic as the access point does. At the same time, no single node can

accurately monitor all the traffic in the air and control the traffic rate of every other node.

Therefore, an effective distributed rate control scheme is needed for the ad hoc mode.
4.4.2 Call Admission Control

In the CAC scheme, three parameters, (TR, TRpeak, len), are used to characterize the

bandwidth requirement of a real-time flow, where TR is the average rate and TRpeak the

peak rate, both in bits/s, and len is the average packet length in bits. For CBR traffic,

TR TRpeak. For VBR traffic, TR < TRpeak. We use the channel utilization cu that a

flow will occupy to describe the bandwidth requirement, and

TR
cu = U(TR) =TR x T,,, (4.8)
len

where U is the mapping function from traffic rate to channel utilization, and T,,, is defined

in equation (4.1) or (4.2). Thus (cu, cupeak) specify a flow's bandwidth requirement, where

cu = U(TR) and cupeak U(TRpeak).

On the side of the coordinator, the total bandwidth occupied by all admitted real-

time flows is recorded in two parameters, i.e., the aggregate (cu, cupeak), denoted by (cuA,

cupeak), which are updated when a real-time flow joins or leaves through the following

admission procedure.









When receiving a real-time connection request from its application layer, a node must

send a request with specified (cu, cupeak) to the coordinator, noting that it wants to establish

a real-time flow. Only after the request is admitted, the node starts to establish the flow

with the intended destination. Otherwise, the node rejects the request and informs the

corresponding application.

Upon receiving a QoS request with parameters (cu, cupeak), the coordinator checks

whether the remainder of the quota BM can accommodate the new real-time flow. Specifi-

cally, it carries out the following:

If cuA + cu < BM and CUpeakA + Cpeak < Bu 1 the coordinator issues the "con-

nection admitted" message, and updates (cuA, cupeakA) accordingly;

Otherwise, the coordinator issues the "connection rejected" message.

Finally, when a real-time flow ends, the source node of the flow should send a "con-

nection terminated" message to the coordinator, and the latter responds with a "termination

confirmed" message and updates (cuA, CUpeakA) accordingly.

Note that real-time packets have highest priority in the outgoing queue, which means

they will always be put on the top of the queue. Meanwhile, all the control messages related

to connection admission and termination are transmitted as best effort traffic; however, they

have higher priority than other ordinary best effort packets, which have the lowest priority.

By doing so, we make sure that these messages do not affect the real-time traffic while

being transmitted promptly.



1 Note that this criterion can provide QoS guarantees for VBR real-time traffic, although
it is conservative if CUpeakA/cuA is much larger than BU/BM. This problem could be
alleviated if we use measured values of cuA or CUpeakA; however, it is well known that
when the number of present real-time flows is small, the CAC must also be conservative in
order not to cause serious QoS degradation [79]. We will further investigate this issue in
our future work.









4.4.3 Rate Control
Rate control in infrastructure mode

We adopt a sliding window smoothing algorithm to estimate the aggregate instanta-

neous bandwidth requirement of the real-time traffic cuA. Let us denote by tient the period

between the (i 1)-th and i-th successful packet transmission or reception at the access

point, and denote by tirel the time consumed by real-time traffic in this period. Apparently,

if the i-th packet is a TCP packet, tirel = 0. Thus we have


cuAri = =i+1-k real i+-kint (4.9)


where k is the sliding window size. Thus the instantaneous available bandwidth for best

effort traffic, denoted by CUbi, is


CUbi BU CUAr (4.10)

If the recent k packets are all TCP packets, then cuAr, = 0 and all the bandwidth will be

allocated to TCP flows. Once a real-time packet which has higher priority in the outgoing

queue is transmitted or received, the rate of TCP flows will be decreased. This algorithm

thus effectively adapts TCP rate to the change of VBR traffic rate. Clearly, if k is small, the

estimation is aggressive in increasing TCP rate; if k is large, the estimation is conservative

[79]. We set k to 10 in our simulation as a tradeoff.

Given cub, the task is to fairly allocate the bandwidth to all the nodes that have the

best effort traffic to transmit. We assume the number of nodes that are the sources of

downstream flows is nd, and the number of nodes that are the sources of upstream flows

is n,. Obviously, the access point knows both nd and n,. Thus the traffic rate for the

best effort traffic allocated to the access point TRba and that allocated to each mobile node

TRbm are as follows.
TRba U-1(CUb X nd/(nl( + nd))
(4.11)
TRb, U-'(cub/(n,, + nd))









where U-1 is the inverse function of U defined in Equation (4.8).

This rate allocation TRba immediately takes effect at the access point. And the rate

allocation TRbm is piggybacked to each mobile node by using the MAC layer ACK frame

for each best effort packet from the node. In this way, the mobile node can immediately

adjust the transmission rate of its own best effort traffic. Two bytes need to be added in the

ACK frame to indicate TRbm with a unit of RD x 2-16, where RD is the bit rate for the

MAC layer DATA packets.

Note that the above fair allocation algorithm is only one choice for rate control. De-

pending on traffic patterns, other allocation algorithms can also be used, since the access

point can monitor the instantaneous rate of each best effort flows from/to each mobile node.

For instance, it is easy to design an algorithm that allocates different rate to different flows

by modifying Equation (4.11).
Rate control in ad hoc mode

We propose a novel, simple and effective rate control scheme for the best effort traffic

at each node. In this scheme, each node needs to monitor the channel busyness ratio Rb

during a period of Trb. Let us denote by Rbr the contribution from real-time traffic to Rb,

and denote by TRb the traffic rate of best effort traffic at the node under consideration, with

the initial value of TRb being conservatively set, say one packet per second. The node thus

adjusts TRb after each Trb according to the following:

Rbt br
Rb Rbr
TRb Tlbld Rb Rb (4.12)

where TRb~,, and TRbol are the value of TRb after and before the adjustment, and Rbt

is a threshold of channel busyness ratio and is set to 95' x Bu. Two points are noted

on Equation (4.12). First, we see that the node increases the rate of best effort traffic if

Rb < Rbt and decreases the rate otherwise. Second, if all the nodes adjust the rate of its

own best effort traffic according to Equation (4.12), the total best effort traffic rate will be


TRb_ TRb ,, x Rbt Rb _U- 1(Rbt Rb), (4.13)
Rb Rb(









where Y TRbod U-1(Rb Rbr) is due to the fact that R, / Rb as shown in Equation

(4.7) and Rb Rb, is the contribution from the total best effort traffic to Rb. Thus after one

control interval Trb, the channel utilization will approximately amount to Rbt.

Apparently this scheme depends on the estimation of Rbr. Larger estimate of Rbr

results in larger increase in traffic rate when Rbt > Rb and larger decrease in traffic rate

when Rbt < Rb. On the contrary, smaller estimate of Rbr results in smaller increase in

traffic rate when Rbt > Rb and smaller decrease in traffic rate when Rbt < Rb. To avoid

overloading the wireless LAN and protect the QoS level of admitted real-time traffic, a

conservatively increasing and aggressively decreasing law is desired for controlling the

best effort traffic rate. This is especially preferred given the fact that an accurate estimate

of Rbr is not available. These considerations have led us to the following scheme to estimate

Rbr.

Each mobile node needs to monitor all the traffic in the air. Note that, to be consistent

with the original 802.11 protocol, our scheme only requires mobile nodes to decode the

MAC header part, as the original 802.11 does in the NAV procedure, instead of the whole

packet it hears. For the purpose of differentiating real-time packets from best effort packets,

one reserved bit in the subtype field of the MAC header is used. Therefore, the observed

channel busyness ratio comprises three pieces of contribution: the contribution from best

effort traffic with a decodable MAC header Rbl, that from real-time traffic with a decodable

MAC header Rb2, and that of all the traffic with an undecodable MAC header Rb3 due to

collision. So we give an upper bound, a lower bound, and an approximation for Rbr as

follows:
Rb2 < Rbr < Rb2 + Rb3
(4.14)
Rbr ~ Rb2 X (1 + Rb )= Rbr
Rbl+Rb2 Rbl+Rb2 -
where we assume Rb3 is composed of real-time traffic and best effort traffic according to

the ratio of Rbr/ b.









To enforce a conservatively increasing and aggressively decreasing law, we thus set

Rbr as follows:

Rb T Rb2, if Rb < Rbt; (4.15)
Rbr= < (4.15)
Rb2+ Rb3, if Rb > Rbt-
We also need to determine the control interval Trb distributedly. To be responsive to

the change of the channel busyness ratio observed in the air, the rate is adjusted at each

time instant when a node successfully transmits a best effort packet. Thus Trb is set to the

interval between two successive best effort packets that are successfully transmitted. Note

that even when such an interval is short and thus no real-time traffic is observed in it, i.e.,

Rbr = 0, the rate of best effort traffic can at most be increased to U -1(Rbt). At that time,

the collision probability is still very small according to previous analysis, so the real-time

packets later on can be quickly transmitted, which will in turn lower the best effort traffic

rate.
4.5 Performance Evaluation of CARC

We have implemented the CARC scheme in ns-2 simulator [106]. In this section, we

evaluate its effectiveness in an 802.11 wireless LAN.
4.5.1 Simulation Configuration

An 802.11 based wireless LAN with 100 mobile nodes is simulated. In all simulations,

channel rate is 2 Mb/s and simulation time is 120 seconds. The queue length at each node

is 100 packets. The IEEE 802.11 system parameters are summarized in Table 4-1

To model multimedia traffic, three different classes of traffic are considered:

Voice Traffic (VBR): The voice traffic is modeled as VBR using an on/off source with

exponentially distributed on and off periods of 300 ms average each. Traffic is generated

during the on periods at a rate of 32 kb/s with a packet size of 160 bytes, thus the inter-

packet time is 40 ms.

Video Traffic (CBR): The video traffic is modeled as CBR traffic with a rate of 64 kb/s

with a packet size of 1000 bytes, thus the inter-packet time is 125 ms.









Data Traffic Model (UBR): We use the greedy best-effort TCP traffic as the back-

ground data traffic with a packet size of 1000 bytes.

During simulation, the RTS/CTS mechanism is used for video and TCP packets, but

not used for voice packets because of its relatively large overhead. The traffic load is grad-

ually increased, i.e., a new voice, video or TCP flow is periodically added in an interleaved

way, to observe how CARC works and the effect of a newly admitted flow on the per-

formance of previously admitted flows. Specifically, until 95 second, a new voice flow is

added at the time instant of 6 x i second (0 < i < 15). Likewise, a video flow is added

two seconds later and a TCP flow is added 4 seconds later. Furthermore, to simulate the

real scenario where the start of real-time flows are randomly spread over time, the start of

a voice flow is delayed a random period uniformly distributed in [Oms, 40ms], and that of

a video flow delayed a random period uniformly distributed in [Oms, 125ms]. Note that in

the simulation period between [95ms, 120ms], we purposely stop injecting more flows into

the network in order to observe how well CARC performs in a steady state.

Two scenarios shown below are investigated.

Infrastructure Mode: In this case, all flows pass through the access point, whereby

half number of flows are downstream, and another half are upstream. The sources or the

destinations of these flows which are not the access point are randomly chosen from all the

mobile nodes other than the access point.

Ad Hoc Mode: In this case, there is no fixed access point. Therefore, the sources and

the destinations of all flows are randomly chosen from all the mobile nodes. All the other

parameters are the same as those in the infrastructure mode.
4.5.2 Simulation Results

From the simulation results, we find there are a total of 12 voice flows and 11 video

flows admitted at 66 second; and no more voice or video flows are admitted thereafter. The

number of TCP flows increases by one every 6 s until 95 second. After 95 s, as expected,

there is no change in the number of flows.









As can be calculated using Equation (4.8), each voice flow contributes 0.0347 to the

channel busyness ratio Rb, and each video flow 0.0! ; :' by noticing that each packet is

added a 20 bytes IP header in ns-2. Thus after 12 voice and 11 video flows are admitted,

the portion of Rb that accounts for the voice flows is 0 ~ 0.38, with a mean of 0.19, and

the portion that accounts for the video connections is 0.52. Thus U(TRA) = 0.71, and

U(TRApeak) = 0.90. Thereafter, the admission control mechanism starts to reject future

real-time flows.
Infrastructure mode

Fig. 4-3(a) shows the throughput for the three traffic types throughout the simulation.

At the beginning, the TCP traffic has high throughput; then as more real-time flows are

admitted, it gradually drops as a result of rate control. Because we set an upper bound

BM for real-time traffic, it can be observed that when the traffic load becomes heavy, TCP

traffic, as desired, is not completely starved. Because TCP traffic is allowed to use any

available channel capacity left by real-time traffic, the total channel throughput, namely

the sum of the throughput due to different types of traffic, always remains steadily high.

Note that the throughput for the TCP traffic does not include the contribution from TCP

ACK packets, even though they also consume channel bandwidth to get through. Thus, the

total channel throughput should be somewhat higher than the total throughput as shown in

Fig. 4-3(a).

The end-to-end delay is illustrated in Fig. 4-3(b), in which every point is averaged

over 2 seconds. It can be observed that the delay for real-time traffic is always kept below

20 ms. Initially, as the number of admitted real-time flows increases, the delay increases.

Note that the increase of delay is not due to TCP traffic, but due to the increasing number

of competing real-time flows. Then, the delay oscillates around a stable value. Fig. 4-

3(c) presents the delay distribution for voice and video traffic. More detailed statistics of

delay and delay variation are given in Table 4-2 and Fig. 4-4. As shown in Table 4-2,

the 97 percentile delay value for voice and video is 35.5 ms and 32.2 ms respectively, and