无线定位与测量 Survey

AsurveyofwirelesslocalizationandmeasurementLingKang20170322WhereAreYouFrom?ConfusingLocationDistinctionUsingVirtualMultipathCamouflage•Idea:–manywirelessauthenticationsystemsrelyondistinctivechannelcharacteristic–WecanmakeachannelcharacteristiccamouflageusingUSRPdeviceTagoram:Real-TimeTrackingofMobileRFIDTagstoHighPrecisionUsingCOTSDevicesTagoram:Real-TimeTrackingofMobileRFIDTagstoHighPrecisionUsingCOTSDevices•IdeaMaximum-likelihoodestimationJigsaw:IndoorFloorPlanReconstructionviaMobileCrowdsensingIdea:combineimageandmobileinertialdatatoconstructindoorfloorplanLuxapose:IndoorPositioningwithMobilePhonesandVisibleLight•indoorpositioningwithunmodifiedsmartphonesandslightly-modifiedcommercialLEDluminaires.ExperiencingandHandlingtheDiversityinDataDensityandEnvironmentalLocalityinanIndoorPositioningService•Akeyinsightsuggestedbythisstudyisthatdifferentlocalizationapproachesarepreferablyappliedunderdifferentconditions(datadensity,environment)Travi-Navi:Self-deployableIndoorNavigationSystem•Travi-Navi–avision-guidednavigationsystemthatenablesaself-motivatedusertoeasilybootstrapanddeployindoornavigationservices,withoutcomprehensiveindoorlocalizationsystemsoreventheavailabilityoffloormaps.•Related:Mobicom2015[inertialsensors]Last-MileNavigationUsingSmartphonesAccurateIndoorLocalizationWithZeroStart-upCost•Twistatablet,suchthatitsantennacanberegardasacircularantennaarray.(SARinradar)WeCanHearYouwithWi-Fi!•WiHearexploitstheradiometriccharacteristicsofmouthmovementstoanalyzemicro-motioninanon-invasiveanddevice-freemanner.UseItFree:InstantlyKnowingYourPhoneAttitudeObject:accuratemobilephoneattitudeCurrent:kalman-basedsenserfusionIdea:opportunisticcalibrationaccordingtoconcordanceofthreetypesofsensors(gyroscope,compass,accelerometer)E-eyes:Device-freeLocation-orientedActivityIdentificationUsingFine-grainedWiFiSignaturesUnderstandingandModelingofWiFiSignalBasedHumanActivityRecognitionWiDraw:EnablingHands-freeDrawingintheAironCommodityWiFiDevices•Devicefreehandgesturerecognition•Upto25WiFitransmitters•Meanwordrecognitionaccuracyof91%KeystrokeRecognitionUsingWiFiSignals•96.4%recognitionaccuracyforclassifyingsinglekeysReusing60GHzRadiosforMobileRadarImaging•60GHzimagingsystem–Objectorientation/position–Objectlength/curvature/materials•AnRSSbasedmethodReusing60GHzRadiosforMobileRadarImaging•60GHzimagingsystem–Objectorientation/position–Objectlength/curvature/materials•Canwedothiswithultrasound?mTrack:High-PrecisionPassiveTrackingUsingMillimeterWaveRadios•60Ghzsingle-carrierphasebasedtrackingAcousticEavesdroppingthroughWirelessVibrometry•Idea:voicefromspeakerissomekindofamodulationtowirelesssignal•Device:WARPsoftwaredefinedradioSnoopingKeystrokeswithmm-levelAudioRangingonaSinglePhoneSnoopingKeystrokeswithmm-levelAudioRangingonaSinglePhoneMoLe:MotionLeaksthroughSmartwatchSensors•Challenges–Lackrighthanddata–Inferwhichfingerexecutedkey-press–Differentuserkey-presshabit–Dependencybetweenconsecutivekey-press•Opportunities–2Dkeyboardplane–Uniformwatchorientation–DictionarymodelHumanSensingUsingVisibleLightCommunication•FirstuserskeletonreconstructionbasedonVLC•Idea:shadows.Anyopaqueobject(e.g.,humanbody)obstructsalightbeam,resultinginasilhouettebehindtheobject.•Result:reconstructsthe3Duserskeletonat60Hzinrealtimewith10◦meanangularerrorforfivebodyjoints.HumanSensingUsingVisibleLightCommunication•Challenges:–ShadowacquisitiondifferentfrequencyLED–Shadow-basedskeletonreconstructionwithlowphotodiodedensitybodysegmentsmatchandkalmanfiltersOkuliExtendingMobileInteractionThroughNear-FieldVisibleLightSensing•Okuli:model-drivenframeworkforfingerpositioningbasedonVLCOkuliExtendingMobileInteractionThroughNear-FieldVisibleLightSensing•Challenges–irregular-shapedfinger,togetherwithirregularLED/PDangulargainpatternsgroomsthelightbyreshapingtheLED/PD’sfield-of-view(FoV)withashroud–interferencefromambientlightsourcetakesadvantageoftheflickeringcapabilityofVLCtransceiverstoperformcommonmoderejectionSeeThroughWallswithCOTSRFIDSystem!•Tadar:SeethroughwallRFIDsystem•Result:–medianerrorsof7.8cmand20cminXandYdimensions–8’’concretewalls,5’’hollowwalls,and1.5′′woodendoors–4m*6mLast-MileNavigationUsingSmartphones•Idea:–Aleaderrecordits(sensor)trace,followerjustfollowit.•FOLLOWME,a“last-mile”navigationsystem–infrastructure-freesolutionusingonlysensorsoncommoditysmartphones–95%ofspatialerrorsduringnavigationwithin2mEchoTag:AccurateInfrastructure-FreeIndoorLocationTaggingwithSmartphones•Idea:differentlocalizationhasdifferentechochannelcharacteristics.ToneTrack:LeveragingFrequency-AgileRadiosforTime-BasedIndoorWirelessLocalizationGyrointheAir:Tracking3DOrientationofBatterylessInternet-of-Things•Tagyro:orientationtrackingbyRFIDtags•Basicidea:anarrayoftagswillenduredifferentPDOA(phasedifferenceofarrival)underdifferentialrotatingangles,combinemulti-arraytoachieve3d-orientationGyrointheAir:Tracking3DOrientationofBatterylessInternet-of-Things•Tagyro:orientationtrackingbyRFIDtags•Challenge:–frequency-hoppingincommercialreaders–couplingeffect–blinddirectionsCAT:High-P