资源描述：
《eigenvalue based spectrum sensing》由会员上传分享，免费在线阅读，更多相关内容在学术论文-天天文库。
1、EigenvaluebasedSpectrumSensingAlgorithmsforCognitiveRadio∗YonghongZeng,SeniorMember,IEEE,andYing-ChangLiang,SeniorMember,IEEEInstituteforInfocommResearch,A*STAR,Singapore.November23,2009Abstracttodevelopastandardforwirelessregionalareanetworks(WRAN)basedoncognitiveradiotech
2、nology[5].WRANSpectrumsensingisafundamentalcomponentisacogni-systemswilloperateonunusedVHF/UHFbandsthataretiveradio.Inthispaper,weproposenewsensingmethodsoriginallyallocatedforTVbroadcastingservicesandotherbasedontheeigenvaluesofthecovariancematrixofsig-servicessuchaswirele
3、ssmicrophone,whicharecalledpri-nalsreceivedatthesecondaryusers.Inparticular,twomaryusers.Inordertoavoidinterferingwiththeprimarysensingalgorithmsaresuggested,oneisbasedonthera-services,aWRANsystemisrequiredtoperiodicallydetecttioofthemaximumeigenvaluetominimumeigenvalue;the
4、ifthereareactiveprimaryusersaroundthatregion.otherisbasedontheratiooftheaverageeigenvaluetomini-Asdiscussedabove,spectrumsensingisafundamentalmumeigenvalue.Usingsomelatestrandommatrixtheoriescomponentinacognitiveradio.Therearehoweverseveral(RMT),wequantifythedistributionsof
5、theseratiosandfactorswhichmakethesensingproblemdifficulttosolve.derivetheprobabilitiesoffalsealarmandprobabilitiesofFirst,thesignal-to-noiseratio(SNR)oftheprimaryusersdetectionfortheproposedalgorithms.Wealsofindthereceivedatthesecondaryreceiversmaybeverylow.Forex-thresholdsoft
6、hemethodsforagivenprobabilityoffalseample,inWRAN,thetargetdetectionSNRlevelatworstalarm.Theproposedmethodsovercomethenoiseuncer-caseis−20dB.Secondly,fadingandtimedispersionofthetaintyproblem,andcanevenperformbetterthantheidealwirelesschannelmaycomplicatethesensingproblem.In
7、energydetectionwhenthesignalstobedetectedarehighlyparticular,fadingwillcausethereceivedsignalpowerfluc-correlated.Themethodscanbeusedforvarioussignaltuatingdramatically,whileunknowntimedispersedchanneldetectionapplicationswithoutrequiringtheknowledgeofwillcausecoherentdetect
8、ionunreliable[6,7,8].Thirdly,signal,channelandnoisepower.Simulationsbasedonran-the
