Abstract
The problem of joint sparsity pattern recovery with a common sparse support based on multiple measurement vectors (MMV) arises in distributed networks including cooperative cognitive radio and sensor networks. Since these wireless networks operate under resource constraints namely communication bandwidth and processing power constraints, finding efficient solutions to this problem becomes challenging. In this paper, we propose several approaches to overcome this challenge when distributed nodes employ Compressive Sensing (CS) so that measurements at each node are obtained via a lower dimensional random sampling scheme. First, we show that the common sparsity pattern can be reliably recovered under certain conditions when each node forwards its observations to a fusion center via a shared multiple access channel (MAC), whose bandwidth requirement does not depend on the number of sensors. Next, we develop two distributed algorithms based on Orthogonal Matching Pursuit (OMP) to perform joint sparsity pattern recovery in a decentralized manner with minimal computational and communication complexities. We further consider the problem in a binary hypothesis testing framework, where a decentralized version of OMP is used to jointly detect the sparse signal using a fewer number of iterations and increase the number of iterations to estimate the sparsity pattern only if the signal is detected. The efficiency and the effectiveness of the proposed algorithms are illustrated by numerical simulations.

This publication has 24 references indexed in Scilit: