Performance Evaluation of Unitary Measurement Matrix in Compressed Data Gathering for Real-Time Wireless Sensor Network Applications

Abstract

Recently, compressed sensing has emerged as a novel phenomenon of simultaneous sampling and compressing any sparse signal. Wireless sensor network which is resource constrained requires to preserve its energy by various mechanisms. Each sensor node in wireless sensor network records the data in its surrounding which generates number of signals in the entire network. Communication process consumes maximum energy in the network as compared to other in network processes. So, energy may be preserved by reducing data rate in network by using distributed compressed sensing. In this paper, we propose to use a unitary matrix as measurement matrix to perform distributed compressed sensing to exploit both spatial and temporal correlation in sensor network data. The parameters used for measuring performance of the proposed scheme are the percentage by which overall network lifetime increases and the mean squared error in reconstruction of the original signal from compressed signal at the sink.