Distributed localization of an unknown target in NLOS environments

We propose a novel distributed on-line expectation maximization method for passive target localization in wireless sensor networks where multiple sensors cooperate to process the location information obtained from the radio signal transmitted by the target. We consider the scenario where few or no sensors receive line of sight signals. Each sensor computes the time difference of arrival of the target´s signal with respect to a reference node and the angle of arrival. Our algorithm is shown to converge and simulation results suggest that our algorithm outperforms the distributed maximum likelihood estimator.