多稀疏基分簇压缩感知的WSN数据融合方法.pdf

《多稀疏基分簇压缩感知的WSN数据融合方法.pdf》由会员上传分享，免费在线阅读，更多相关内容在行业资料-天天文库。

1、第29卷第3期传感技术学报V01．29NO．32016年3月CHINESEJOURNALOFSENSORSANDACTUATORSMar．2016TheMethodofDataAggregationforWirelessSensorNetworkBasedonClusterCompressedSensingofMulti—SparsityBasisZHULu，LIUYuanyuan，CIBaishan，PANZezhong(SchooloflnformationEngineering，EastChinaJiaotongUniversity，Nanehang330013

2、，China)Abstract：AnoveldatafusionmethodforWSN(WirelessSensorNetwork)basedonclustercompressedsensing(CCS)ofmulti—sparsitybasisispresentedtosolvethecontradictionbetweendataaccuracycollectedandenergyconsumptioninsensornodes．Intheproposedmethod，theimprovedthresholdisadoptedtoselectclusterhea

3、dandformoptimizationclusterfromtherandomdeploymentofsensornodes，andtheBernoullirandommatrixisutilizedtolinearlycompresssensordataintheclusterbyeveryclusterhead，thenthecompressedinformationistransmittedtothesink，SOitreducesdatatransmissionandenergyconsumptionofcommunication，thusimproving

4、thelifetimeofnetwork．Accordingtomonitorsignalsbeingofsparsityinfinitedifferenceandwavelets，thesinkusesOOMPa1．gorithmtoreconstructlinearcompressionprojectioninformationfromthefinitedifferenceandwaveletssparsityba．sisrespectively．Andtheleastsquaremethodisadoptedtogettogetherthetwodifferen

5、treconstructionsignalswhichcanimprovedataaccuracy．SimulationexperimentresultsshowthatthedatafusionmethodofWSNbasedonCCSofmulti-sparsitybasiscanguaranteedataaccuracycollected，andimprovethelifetimeofwholenetworkatthesametime，tosolvethecontradictionbetweendataaccuracycollectedandnetworklif

6、etime．Keywords：wirelesssensornetwork；clustercompressedsensing；datafusionEEACC：7230；7220doi：10．39690．issn．1004—1699．2016．03．019多稀疏基分簇压缩感知的WSN数据融合方法朱路，刘媛媛，慈白山，潘泽中(华东交通大学信息工程学院，南昌330013)摘要：针对传感器节点采集数据精度与能量消耗的矛盾，提出多稀疏基分簇压缩感知的无线传感器网络WSN(WirelessSensorNetwork)数据融合方法。该方法利用改进的阈值对随机部署的传感器节点进行簇首选

7、择继而形成最优簇，簇首采用伯努利随机观测矩阵对簇内节点信号进行线性压缩投影，然后将压缩的信息传送给汇聚节点，减少数据传输即降低通信能耗，从而提高网络的生命周期。根据传感器节点监测信号在有限差分和小波中都具有可压缩特性，汇聚节点在有限差分和小波两个稀疏基的约束下，利用OOMP算法分别对线形压缩投影信息进行重构；并采用最小二乘法融合重构信号，提高数据精度。仿真实验结果表明，多稀疏基分簇压缩感知的WSN数据融合方法在减少数据发送的情况下，能提高整个网络的生命周期，解决采集数据精度与网络生命周期的矛盾。关键词：无线传感器网络；分簇压缩感知；数据融合中图分类号