无线传感器网络中分布式单目标检测技术分析-analysis of distributed single target detection technology in wireless sensor networks.docxVIP

  • 1
  • 0
  • 约5.14万字
  • 约 78页
  • 2018-08-07 发布于上海
  • 举报

无线传感器网络中分布式单目标检测技术分析-analysis of distributed single target detection technology in wireless sensor networks.docx

无线传感器网络中分布式单目标检测技术分析-analysis of distributed single target detection technology in wireless sensor networks

AbstractWirelessSensorNetworks(WSNs)providesthemostdirectservicetopeopleinthenextgenerationnetworkwhichhasattractedmuchattentionintheworld.DistributedtargetdetectiontechnologyisthebasicuseofWSNsanditusesrandomlydeployedsensornodestodetectthetargetsignal,thenlocalsensornodessendtheresultstothefusioncenteraftertheymakethecorrespondingdecisions.Thefusioncentermakesthesystemdecisionbythespecificdecisionfusionruleaccordingtothereceivedjudgmentsfromthelocalsensornodes.Inthisdissertation,wemainlyfocusonthedistributedtargetdetectionofthelocaldecisionalgorithmandsystemdatafusionalgorithmbasedonthestudyingofthewirelesssensornetworks.WeanalysetheBayesandNeyman-Pearson.Andthenweputforwardthemultiplysampledetection,thresholdadjustingalgorithmandthehierarchicaldistributedtargetdetectionalgorithmbasedonclustering.Thecontributionsofthisdissertationaredescribedasfollows:Inthestudyofthedistributedtargetdetectionalgorithmbasedonsinglesample,weproposethealgorithmbasedonmultiplesamples.Thealgorithmabandonsthedisadvantagesduetosinglesamplecausedbyrandomness.Ittakesmultiplysampleoftheobservedsignalintoconsiderationandmakesfulluseofthemtoimprovethedetectionperformanceofthelocalsensornodes.Basedonthedistributedtargetdetection,thispaperproposesanadaptivethresholdadjustingalgorithmaccordingtothecharacteristicsofthecomplexbackgroundnoise.Fortheinitialvalueofanygiventhreshold,wecangettheresultclosedtothetheoreticalvaluethroughtheadaptivethresholdadjustingalgorithmwithouttheformulaofthethreshold.Onthebasisofthepracticalapplicationinwirelesssensornetworks,weproposethehierarchicaldistributedtargetdetectionalgorithmbasedonclusteringthatissuitableformonitoringawiderangeofinterests.Thealgorithmnotonlyconsiderstheattenuationofthetargetsignal,butalsotakesconsiderationoftheRayleighfadingofthelocalsensor’sjudgments.Wededucethefusioncriterionoftheclusterheaderfromtheoriginallikelihoodratetestingfunction.WeverifytheeffectivenessandthepracticeofthefusioncriterionalgorithmbasedonMultiplySample,Atalgorithmand

您可能关注的文档

文档评论(0)

1亿VIP精品文档

相关文档