无线传感网络数据的稀疏修复与压缩传输.pdf

ABSTRACT Inwirelesssensor to atasensor networks，due node，hardware poweroutrage channelorenvironmental allsensor callbe dysfunction,bad conditions，notsamples at this aimtorecoverthe successfully thesis，we gatheredthe．sink．In missingsamples withoutretransmission．Basedonthe sensornetworkeddataandthe spars@of theory of use torecoverthe data． compressedsensing，weoptimizationalgorithms missing First,the estimationis formulatedasa missingsamples problemmathematically 2-D the2一DsensordataCanbe spatial interpolation．Assuming sparselyrepresented a for norm by dictiona巧，asparsity-basedrecovery bysolving approach Zl minimizationis isshownthatthese canbe proposed．It missingsamples reasonably recoveredbasedonthenull ofthe also spacepropertydictionary．TIlispropertypoints outthe to an furtherreduce choose to the dictionary way appropriatesparsifying errors．111esimulationresultsdemonstratethatthe call recovery proposed approach recoverthe data wellandthatit the missingreasonably outperformsweightedaverage methodswhenthedata