WSNs中的分簇式压缩感知

Compressive sensing based on clustering network in WSNs

提出了基于无线传感器网络的分簇式压缩感知算法。首先研究了随机压缩采样技术在硬件资源受限的无线传感器网络节点中的技术解决方案。在此基础上提出了在簇头中生成随机采样序列并分发给簇成员,然后在簇成员中进行低速随机采样,最后在簇头中进行信号重构的分簇式压缩感知算法。声音信号的压缩感知实验表明,压缩感知技术能够在硬件资源受限的节点中实现,且提出的分簇式压缩感知算法在没有信息丢失的压缩采样下大大降低了通信能耗。

This paper puts forward a compressive sensing algorithm based on clustering network in Wireless Sensor Network. Firstly, the technical solution scheme of random compressive sensing technique in the nods of wireless sen- sor network with limited hardware resource is studied. Secondly, a compressive sensing algorithm based on clustering network is proposed, in which the random sampling sequences are created in the cluster head and distributed to the cluster members;then low speed random samplings are carried out in the cluster members. A small amount of meas- urements are needed in each cluster member with its own random sampling sequence;and the measurement results are sent to the cluster head and used to reconstruct the signal. The experiment results of the acoustic signal compress- ing sensing show that the compressive sensing technique can be realized in WSNs nodes with limited hardware re- source, and the proposed compressive sensing algorithm based on clustering network can greatly decrease the data transmission and total enerzv consumption without losing information.