A joint recovery algorithm for multiple sensor signal based on basis pursuit denoise
Zhao Shengying, Kai Xiong, Lina Shang, Cui Zhang, Guangchun Gao · 2015
Considering a sensor network deploying a large number of nodes, a new joint recovery algorithm is proposed for multiple sensor signal. The number of measurements at each sensor can be reduced in the case of efficient sensing by using proposed algorithm. Reducing measurement number of sensor is an efficient approach to decrease the energy consumption and extend lifetime of sensor networks. In this paper, measurement matrix yields distributed measurement structure. We modify the traditional basis pursuit denoise algorithm based on Spectral Projected Gradient(SPG-BPDN) using block sparse structure of signal ensembles measured by sensors, and prune the solution vector of each iterate according to signal structure in process of solving the convex optimization problem. Compare with the similar algorithm. The improved algorithm can obtain the same recovery performance with less number of measurement.