Maximum Likelihood trajectory estimation of a mobile node from RSS measurements

In this paper we present a Maximum Likelihood (ML) trajectory estimation of a mobile node from Received Signal Strength (RSS) measurements. The reference scenario includes a number of nodes in fixed and known positions (anchors) and a target node (blind) in motion whose instantaneous position is unknown. We first consider the dynamic estimation of the channel parameters from anchor-to-anchor measurements, statistically modeled according to the well-known Path-Loss propagation model. Then, we address the ML estimation problem for the position and velocity of the blind node based on a set of blind-to-anchor RSS measurements. We compare also the algorithm with a ML-based single-point localization algorithm, and discuss the applicability of both methods for slowly moving nodes. We present simulation results to assess the accuracy of the proposed solution in terms of localization error and velocity estimation (modulus and angle). The distribution of the localization error on the initial and final point is analyzed and closed-form expressions are derived.