摘要：

无线传感器网络(WSN)具有应用灵活和信息感知有效的特点。压缩传感(CS)将采样与压缩过程进行合并,可以降低信号采样频率,节省存储和传输成本。为将CS理论有效应用到WSN中,提出一种基于时空相关性的块压缩感知全局重构算法BCS-STGR。研究常见测量矩阵的性能并优化WSN的拓扑结构,利用扩散小波对网络进行切分后在每个子网中独立进行数据聚集,最终由选定的中心节点将数据传输给sink接收端。仿真结果表明,BCS-STGR算法的归一化平均绝对误差小于5%,优于传统CS算法和基于时空相关性的分块重构算法。

关键词: 压缩感知, 无线传感器网络, 测量矩阵, 时空相关性, 重构算法

Abstract:

Wireless Sensor Network(WSN) has the characteristics of flexible application and effective information perception.Compressed Sensing(CS) combines sampling and compression process to reduce signal sampling frequency and save storage and transmission cost.In order to effectively apply CS theory to WSN,a block compressed sensing global reconstruction algorithm named BCS-STGR based on spatio-temporal correlation is proposed.The performance of the common measurement matrix is studied,the topology of the WSN is optimized,the network is divided by the diffusion wavelet,and the data is gathered independently in each subnet,and the data is transmitted to the sink receiver by the selected central node.Simulation results show that the normalized mean absolute error of BCS-STGR algorithm is less than 5%, which is much better than the traditional CS algorithm and the block reconstruction algorithm based on spatio-temporal correlation.

Key words: Compressed Sensing(CS), Wireless Sensor Network(WSN), measurement matrix, spatio-temporal correlation, reconstruction algorithm

中图分类号: 

• TP393