Distributed Bayesian Sparse Signal Recovery Algorithm with Minimal Communication Load in Networks

We address the problem of distributed recovery of sparse signals in a resource constrained network. We assume that the network nodes sense a common sparse signal and therefore share an approximately common support. We propose a Bayesian algorithm that performs distributed recovery of the sparse signals and is agnostic to the sparse signal distribution. The algorithm requires nodes in the network to communicate only with their neighbors to estimate the sparse signals and is designed to reduce the communication load between nodes. Simulations have been performed to show that the algorithm requires significantly less communication among the nodes as compared to other algorithms.