07-Data-Collection-through-Vehicular-Sensor-Networ

DataCollectionthroughVehicularSensorNetworksAbstract—NowadaysManycarmanufacturersareplanningtoinstallwirelessconnectivityequipmentintheirvehiclestoenablecommunicationswithroadsidebasestationandalsobetweenvehicles,forthepurposesofsafety,drivingassistance,andentertainment.Onedistinctfeatureisthatvehiclesarehighlymobile,withspeedupto30m/s,thoughtheirmobilitypatternsaremorepredictablethanthoseofnodesinMobileAd-hocNetworks(MANET)duetotheconstraintsimposedbyroad,speedlimits,andcommutinghabits.Therefore,thesenetworksrequirespecificsolutionsandidentifyanovelresearcharea,i.e.,VehicularAd-hocNetworks(VANET).Inthispaper,wefocusonaparticularVSNarchitecture,wheretheadhocnetworkisoperatedbyatelecommunication/serviceprovidertocombinenon-valuableindividualsenseddataandextractfromthemeffectivefeedbacksaboutthesituationoftheroadinageographicalarea.InoperatedVSNs,providerstendtoreducethetrafficloadontheirnetwork,usingthefree-frequencycommunicationmedium(IEEE802.11p,forexample).Todoso,weproposeTCDGP(TreeClusteredDataGatheringProtocol),acrosslayerprotocolbasedonhierarchicalandgeographicaldatacollection,aggregationanddisseminationmechanisms.Weanalyzetheperformancesofoursolutionusingasimulationenvironmentandrealisticmobilitymodels.WedemonstratethefeasibilityofsuchsolutionandshowthatTCDGPofferstheoperatorpreciousinformationwithoutoverloadinghisnetwork.Keywords—VSN,VANET,MANET,dissemination,ITS,datacollection,dataaggregation,hybridarchitecture,operatednetwork.1.INTRODUCTIONOverthepastdecadethenatureofwirelesscommunicationshasevolvedrapidly.Theintroductionof3GandWLANtechnologiesandtherecentstandardizationofWiMaxhavehelpedtorealizethevisionofubiquitousconnectivity.Currently,muchresearcheffortisfocusingonexploitingthisalways-onfeatureforuseinTransportationSystems.TheprimaryobjectiveofITS(IntelligentTransportationSystems)istoimprovetrafficsafety,efficiency,andtravellingcomfort.VSNscanbebuiltontopofVANETbyequippingvehicleswithonboardsensingdevicesasshownintheabovefigure.Here,sensorscangathernotonlysafety-relatedinformation,e.g.,seatoccupation,butalsomorecomplexmultimediadata,e.g.,videodata.Unliketraditionalsensornetworks,VSNsarenotsubjecttomajormemory,processing,storage,andenergylimitations.However,thetypicalscaleofaVSNoverwidegeographicareas(e.g.,millionsofnodes),thevolumeofgenerateddata(e.g.,streamingdata),andmobilityofvehiclesmakeitinfeasibletoadopttraditionalsensornetworksolutionswheresenseddatatendstobesystematicallydeliveredtosinksusingdata-centricprotocolssuchasDirectedDiffusion.Further,themobilityofsensornodesmakesitlessefficienttousemobileagents,orMULEsinstaticsensornetworks,whichpickupdatafromsensorswhenincloserange,bufferit,anddropoffthedatatowiredaccesspoints.Thisprojectmainlyfocusesondevelopingefficientdatagathering,searchingprotocols.Moreover,thegoalincludestest-bedimplementationandmiddlewaresoftwaredevelopment.BesidesDSRC,wecanutilizecellularcommunications(2/3G)viaSmartphone’s.RecentSmartphone’sareequippedwithvarioussensorssuchasGPS,camera,audio,andvideo,andsupportvariouscommunicationsmeanssuchas2/3G,Wi-Fi,andBluetooth.Bluetoothenablesustoconnectotherexternalsensorsviaawirelessdataacquisitionboard.Theparticularimportanceof2/3Gconnectionisthatitgivesanalways-onInternetconnection,whichmakesdataaccessandretrievalamenable.Inthiswork,wefocusonthemaincomponentinIntelligentTransportationSystems,whichisvehicularcommunication.Indeed,manycarmanufacturersareinstallingwirelessconnectivityequipmentsintheirvehiclestoenablecommunicationbetweenvehiclesandalsowiththeinfrastructure.VehicularSensorNetworks(VSNs)canbebuiltontopofthesevehicularnetworksbyequippingvehicleswithonboardsensingdevices.Insuchcase,sensorscangatherasetofinformationlikevideodata,speed,localization,acceleration,temperature,seatoccupation,etc.Comparedtotraditionalsensornetworks,thisrecentlyemergedsensornetworkisnotrestrictedbythepowersupplyandthestoragespace.However,thetypicalscaleofaVSNoverwidegeographicareas(e.g.,millionsofnodes),thevolumeofgenerateddata(e.g.,streamingdata),andmobilityofvehiclesmakeitinfeasibletoadopttraditionalsensornetworksolutionswheresenseddatatendstobesystematicallydeliveredtosinksusingdata-centricprotocolssuchasDirectedDiffusion[1].Undersuchenvironment,aneffectiveandefficientarchitecturefordatacollectionanddataexchangeismoreimportant.Thisworkdealswithsuchasystemframeworkconsistingofmobilevehicularsensorsandroad-side-unitsoperatedbyanoperatororaserviceprovider(WiMaxaccesspoint,2.5/3Gbasestation).Road-side-unitsaredistributedovertheroadforcollectingdatafrommobilevehicularsensorspassingby.Whilemobilesensorsonvehiclessenseandsendtheinformationtoroadsideunits.Inourstudy,theVSNwillbeusedbyaninfrastructurewirelessnetworkowner(telecomorInternetprovider,forexample)togatheruselessindividualinformation(thatcannotbeusedtodeduceaglobalviewoftheroadtraffic)fromeachvehicleandtoaggregatetheminsideanadhocwirelessnetworkusingfreefrequency(IEEE802.11,forexample),togetaglobalviewofthestateoftheroadinageographicalareaataspecifictime,ortousetheseinformationasadatabase