Sparse Target Localization in RF Sensor Networks using Compressed Sensing

In this paper, we propose a greedy sparse recovery algorithm for target localization with RF sensor networks. The target spatial domain is discretized by grid pixels. When the network area consists only of several targets, the target localization is a sparsity-seeking problem such that the Compressed Sensing (CS) framework can be applied. We cast the target localization as a CS problem and solve it by the proposed sparse recovery algorithm, named the Residual Minimization Pursuit (RMP). The experimental studies are presented to demonstrate that the RMP offers an attractive alternative to OMP for sparse signal recovery, in addition, it is more favorable than non-CS based methods for target localization.