Compressive data gathering in wireless sensor networks via group sparse regularization

Abstract

In wireless sensor networks (WSNs), how to gather the sensory data efficiently is a major challenge to deploy effective sensor systems. The compressive sensing (CS) provides an improved data acquisition method in WSNs by exploiting a priori data sparsity information. Moreover, the concept of group sparsity has been evolved into signal recovery, which can further enhance the recoverability. In this paper, we extend the compressive sensing framework to group sparsity optimization for incorporating correlation between sensory data. Furthermore, for using the prior information of the signal, support set is employed to further improve the sparsity-measurements tradeoff of the traditional CS recovery methods. The truncated l <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sub> , <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">i</sub> -norm minimization based on the iterative support detection is proposed for the group sparse signals recovery in WSNs. Experimental results demonstrate the accuracy and efficiency of the proposed algorithm using real datasets.