The Application of Wavelet Threshold on Compressive Sensing in Wireless Sensor Networks

Abstract

Compressive sensing (CS) is a novel framework which exploits both the sparsity and the intra-correlation of the signal in structural health monitoring (SHM) based on wireless sensor networks (WSNs). It contains sparse signal representation, the measurement matrix selection and the reconstruction algorithm. The SHM signal is recovered by M measurements following the restricted isometry constant (RIC). However, the signal should be denoised before reconstruction. This paper discusses two wavelet noise reduction methods, soft threshold and hard threshold method, and verifies the performance of different methods for SHM signal reconstruction. Experimental results show that wavelet hard threshold method has much better effect on SHM sparse signal reconstruction than soft threshold method. Meanwhile, we can get a more accurate corresponding relation of RIC that is