基于粒子濾波的MIMO時變信道估計.pdf

1、北京交通大學(xué)碩士學(xué)位論文基于粒子濾波的MIMO時變信道估計姓名：劉艷芳申請學(xué)位級別：碩士專業(yè)：交通信息工程及控制指導(dǎo)教師：張三同20090701jE立鑾通塞堂亟堂位詮塞△旦S!B△￡!ABSTRACTABSTRACT：Withthedevelopmentofmobilecommunication，thedemandsforthelimitedfrequencyresourceareboomingThemajorchallengeinto

2、day’SwirelesscommunicationishowtoservetheexplosivelyincreasingdemandofmultimediaservicewithinthelimitedbandwidthAmongthoseemergingtechnologies，multipleinputmultipleoutput(MIMO)communicationarchitecturehasshoweditsinfinit

3、epotentialtoimprovespectralefficiencydramaticallyBydeployingmultipleantennaelementsatboth仃ansmitterandinevitablepromisingapproachtoexploitreceiverMIMOhasbeenregardedasanmaximumspatialresourcesamongmanyemergingtechnologie

4、sItCanenormouslyincreasethewirelesscommunicationsystemcapacityandthedataratewithoutanincreaseinbandwidthandtransmissionpower，therebyprovidingmuchhigherbandwidthefficiencythanthetraditionalsingleantennasystemToachieveaMIM

5、Osystem，thereceivernotonlyneedthelowcomplexityandlligh—performancedetectionalgorithmbutalsoneedthechannelestimationofMIMOMIMOchannelhasfarmoreparametersthanthetraditionalsingleinputsingleoutput(SISO)channel，SOchannelesti

6、mationisverycomplexityanditisdifficulttoachieveforthereceiverMonteCarloalgorithmbasedonrecursivepropagationAparticlesetwhichisrandomlysampledfromprobabilityfunctionandhascorrespondingweights，isintroducedtoapproachthepost

7、eriordistributionThereforeitcallhandlenonlinearandnonGaussianproblemswithoutanylimitsItisthemainmethodtodealwithnonlinearnonGaussianproblemBasedontheoutcomeofpreviousstudythethesisfocusesonparticlefilteralgorithmandappli

8、edittomultiple—inputmultipleoutputchannelestimationIIlthisthesis，atfirst，basedonthewirelesschannel，ithasmainlymaderesearchontherelatedtheoriesaboutthreemainaspects：channelmodels，channelcapacitychannelestimationandimpleme

9、ntationalgorithmsAnditanalyzedthecharacteristicsofMIMOchannel，establishedthemathematicalmodelofMIMOchannel，introducedthetraditionalchannelestimationalgorithmofMIMOBasedonthetheoryofBayesianestimationandMonteCarlomethods，