传感技术学报2016，Vol.29Issue(3)：417-422,6.DOI:10.3969/j.issn.1004-1699.2016.03.019
多稀疏基分簇压缩感知的WSN数据融合方法
The Method of Data Aggregation for Wireless Sensor Network Based on Cluster Compressed Sensing of Multi-Sparsity Basis
摘要
Abstract
A novel data fusion method for WSN(Wireless Sensor Network)based on cluster compressed sensing (CCS)of multi-sparsity basis is presented to solve the contradiction between data accuracy collected and energy consumption in sensor nodes. In the proposed method,the improved threshold is adopted to select cluster head and form optimization cluster from the random deployment of sensor nodes,and the Bernoulli random matrix is utilized to linearly compress sensor data in the cluster by every cluster head,then the compressed information is transmitted to the sink,so it reduces data transmission and energy consumption of communication,thus improving the lifetime of network. According to monitor signals being of sparsity in finite difference and wavelets,the sink uses OOMP al⁃gorithm to reconstruct linear compression projection information from the finite difference and wavelets sparsity ba⁃sis respectively. And the least square method is adopted to get together the two different reconstruction signals which can improve data accuracy. Simulation experiment results show that the data fusion method of WSN based on CCS of multi-sparsity basis can guarantee data accuracy collected,and improve the lifetime of whole network at the same time,to solve the contradiction between data accuracy collected and network lifetime.关键词
无线传感器网络/分簇压缩感知/数据融合Key words
wireless sensor network/cluster compressed sensing/data fusion分类
信息技术与安全科学引用本文复制引用
朱路,刘媛媛,慈白山,潘泽中..多稀疏基分簇压缩感知的WSN数据融合方法[J].传感技术学报,2016,29(3):417-422,6.基金项目
国家自然科学基金项目（31101081，61162015）；江西省科技支撑项目 （）
