MetaSensing: Reconfigurable Intelligent Surface Assisted ...

1 metasensing : Reconfigurable Intelligent Surface Assisted RF sensing and Localization !"#$%!"#$%&'!"#$%&'"(&)*+),-".&%/.$-)$(0 )1*'#2&"%),(3/(""%/(34(/5"%6/&7)*+)8*26& *(9)8*26 9)4<=>&&#?@@ BoyaoDi, Hongliang Zhang, LingyangSong, JingzhiHu, HaoboZhangTable of Contents! Background 6G Communications and Requirements RIS Basics and Potential Applications RIS-aided RF sensing and Localization Case 1: Posture Recognition Case 2: RF 3D Shape sensing Case 3: Ubiquitous Localization Potential Future Directions ConclusionsMoving Towards 6G: Emerging Use Cases"VR/ARInternet-of-ThingsAR for surgeryVR for educationAuto-manufacturingE-healthIntel ligenceEnvironment sensingSmart home#General 6G KPI TargetsGreatersensing capabilityHigher positioning accuracy5 GmMTCURLLCeMBB6 GEnergy efficiencyConnectivitySensing accuracyPeak data rateLatencySecurityConflict between simplicity (comfort) and high sensing accuracy$mmWaveRadar6G Challenges.

2 sensing Efficiency WiFibased RF sensing Requires the cooperation of multiple WiFiaccess points to achieve high sensing accuracy mmWaveRadar High hardware cost makes it hard for mass deploymentWiFibased RF SensingRadarSolutions: Meta-Material aided SensingExpectation on a new technology Low cost in manufacture Easy and flexible deployment Compatible with 6G demands on sensing and localizationReconfigurable Meta- surfaces Implemented by metamaterial Cost efficient in manufacture and deployment Control and customize favorable radio environments Provide high accuracy contact/contactless sensing with wireless data gathering So-called Reconfigurable Intelligent Surface ,(RIS) or Intelligent Reflecting Surface (IRS)%RISRF antennasIntroduction of MetamaterialNatural Materials Limited EM Wave Control Capability The dielectricpermittivity, ,and magnetic conductivity, , of materials determine the capability of controlling EM waves( reflection, refraction)

3 Limited possibilities of atom arrangement of natural materialslead to limited available values of and , and thus limited capability to control EM waves& Common materials( , water and glass)Magnetic materialsat low frequencyMetal materialsat light frequencyRare in natureAtom arrangement has limited possibilitiesMetamaterials: Powerful EM Wave Control Capability Metamaterials are artificial structuresthat are non-existent in nature and can have arbitrary pair of ( , ) Two technology fields studying metamaterials Opticsand MicrowaveHistoryof Metamaterial DevelopmentVeselago. Concept of Left-handed material <0, <0 Negative refractionPendry. Realize - and - - : periodic array of metallic rods - : periodic array of split ringSievenpiper. Proposal of meta- Surface Two-dimensional Simplify design and manufacturingSievenpiper.

4 Programmable metasurface Varactor 360!reflection phase tuning19681996 & 199919992001D. R. Smith. Experimental verification Left-handed material20012006 Pendry, et al. Transformation optics Design metamaterial with any and Enabling flexible control of EM wave'Historyof Metamaterial Development20112014201920192020H. Kamoda, et al. Reconfigurable large reflectarraywith PIN diodes Easy to control Millimeter waveT. Cui, et al. Programmable metasurfacewith PIN diodes Simplify the design Digital codingM. D. Renzo, et al. Proposal of Reconfigurable Intelligent surfaces Focus on reflection Extensive applications in wireless networksNTT Docomo. Protype of metamaterial reflector 10x increase in data rate(S. Zhang, et al. Proposal of Intelligent omni- Surface Enabling dual function of reflection and transmissionReconfigurable Intelligent surfaces (RIS)) Outer layer: A 2D-array of RIS elements; directly interact with incident signals.

5 Middle layer: A copper plate; prevent the signal energy leakage. Inner layer: A printed circuit; connect the RIS elements to the RIS controllerCopper backplaneControl circuit boardRIS element Low-cost sub-wavelength programmablemetamaterialparticle. Reflectincident RF signals andimpose a controablle phase shift Working frequency: from sub-6 GHz to THzAn ultra-thinmetasurface composed of multiple layersExample of a programmable metamaterial particlePIN diodeWorking Principle for Wireless Communications!*RISworksasabeamformer Signals can be reflected or transmitted Phase shift of the radiation is controlled by PIN diodes bias voltages (ON/OFF of the diode) Programming the ON/OFF of all diodes collectively realize different beamforming modesAdvantage Cost efficiency: Analog beamforming, no extra RF equipmentneeded for demodulation & modulation Energy Efficiency: No extra RF signals generation, energy savingPolarizationScatteringFocusing0000 00111111010101001011ON/OFF coding of diodesReflected beamIncident: !

6 "#$%Reflect: !"#$%&'RIS elementTransmit: !"#$%&'Signal ReflectionModel!!Model of reflected signal on RIS = () [0,1]: reflection amplitude =0: absorbed =1: fully reflected [0,2 ]: phase shiftbetween incident and reflected In practical systems, available phase shifts of an RIS element are discrete, due to limited number of PIN diodes (K PIN diodes 2*phase shifts). The parameters of an RIS element1are carefully designed so that the phase shifts have uniform : , shape of the metal patch and type of the PIN diodes , Incident EM waveEmitted EM radiationfrom inducedcurrentMechanisminducedcurrent ,etal, ReconfigurableIntelligentSurfacesassiste dCommunicationswithLimitedPhaseShifts:Ho wManyPhaseShiftsAreEnough? IEEET ransactionsonVehicleTechnology, , , , Model!"RicianModel User-RIS-BSlinksactasthedominantLoScompo nent AllotherpathscontributestheNLoS Productofdistancepathloss ReceivedsignalNLoSLoSRatio of LoSto NLoSReflection coefficientChannel gainnoise !

7 ,#, LoS component) !,#, NLoS componentApplications: Radio Frequency sensing !#Indoor Localization and Recognition Enhance remote RF sensing by customize radio environments. Enable high accuracyindoor human and object localizationand recognitionCustomized radio environement for sensinghuman postureCustomize signal beams for scanning human location!$Prototype of Metasurface Size of metasurface:45 57 cm3, total 640 metamaterial particles Totalnumberofpossiblephaseshifts:4 2ofthemareused,andhavephaseshiftswithint erval ParticlePhotoPIN #1,2,3 Choke inductanceSubstrateVia holeCopper PatchMetal layerStructure* Photo shows the actual metasurface prototype used as the testbed in PKU of Contents!% Background 6G Communications and Requirements RIS Basics and Potential Applications RIS-aided RF sensing and Localization Case 1: Posture Recognition Case 2: RF 3D sensing Case 3: Ubiquitous Localization Potential Future Directions ConclusionsRFsensing Livingenvironmentiscoveredseamlesslybywi relesssignals Ubiquitoussignalsprovidethefoundationfor RFsensingBackground!

8 &Received signals change due to change of target Principles: SensingtargetsbetweenapairofRFTxandRximp acttheRFchannel. of cellular signalsApplications!' Advantages: Noneedsforthecontactorline-of-sightviewo fthesensingtargetsSecuritySmart SpaceSafetyTheft DetectionTheft detectionFall DetectionElderly CareInteractionEmergency AlarmTechniques Review!( ActiveMethods: WiFiSensing: UtilizetheimpactofthetargetsonWiFisignal s Variousmetrics:signalstrength,phase,dopp lerandsoon mmWaveRadar: UtilizethedirectionalbeamsinmmWavecommun ications Receiverscandetectreflectedsignalsfromta rgets PassiveMethod:RIS-aidedRFsensing Limitations:sensingaccuracyislimitedbych annelconditionsSolutionGoals and Challenges!)Goals ImplementpracticalRIS-aidedRFsensingsyst emforhumanandobjectlocalizationandrecogn ition AchievehighsensingaccuracyChallenges Designpracticalsensingprotocolstocoordin atetheRISandtheRFtransceiver.

9 Searchtheoptimalphaseshiftselectionforth eRISelementsinalargefeasibleregion. RF Posture sensing : Design, Optimization, and ImplementationCase Study I: RIS aided Posture Recognition"* ,etal, ReconfigurableIntelligentSurfacesbasedRF sensing :Design,Optimization,andImplement ation, IEEEJ ournalofSelectedAreasinCommunications, , , , RIScancontrolthewirelessenvironment,whic hcanprovidefavorablewirelessenvironmentf orRFsensing. Applicationinhumanposturerecognition: Recognizedifferenthumanposturesautomatic allyMotivation"!Challenges RISconfigurationdesign:HowdoesRIScontrol thewirelessenvironment Thediscretephaseshiftsofamassivenumberof RISelementsneedtobedetermined. Decisionfunctiondesign:HowdoesRxjudgehum anposture RFchannelsinvolveanRISandapracticalareha rdtoanalyze,whichmakestherelationshipbet weenRxsignalsandhumanpostureinexplicit.

10 Moreover,RISconfigurationanddecisionfunc tionarecoupled.""Model DescriptionSystemStructure Transmitter:Adirectionalantennawhichispo intedtowardstheRIS Receiver:Anomni-directionalverticalanten nabelowtheRIS Human!Spacereflectionvectorcarriestheinf ormationofpostures. Multi-pathcomponent: Environmentscattering LoScomponent: Transmitter ReceiverRISelement Reflectiondominatedcomponents Transmitter RIS Human Receiver"#Periodic Configuring ProtocolRecognitionPeriod: Contains frames,duringwhichthehumanpostureisfixed Receivedsignalsduringarecognitionperioda reusedforrecognitionFrameConfiguration: EachgroupofRISelementssequentiallychange sfromState1toNa. Constituted by the durations that each group stays in the +statesDifferentstatescorrespond to :Minimizethefalserecognitioncost(Bayesia n)Problem Formulation"$min , &'( , )=3(,("Pr(pos() ( , )@ [Pr( |pos() ("( )]ProblemDecomposition: Decomposing(P1)intotheframeconfiguration optimizationandthedecisionfunctionoptimi zation.)))))