Radio Source Localization Using Received Signal Strength in a Multipath Environment

We consider the problem of localizing a radio emitter in a wireless network using RSS measured by a set of known network nodes in a multipath environment. While the RSS of a wireless signal can be conveniently accessed, using it to estimate location is nontrivial in the presence of multipath. We propose a HMM model within a Bayesian learning framework for processing RSS data in the localization process to deal with RSS fluctuations induced by multipath interference. To address the uncertainty of emitter dynamics, a semi-Markov model is also adopted to model the duration time of the emitter sojourn in a state. We compare the performance of the HMM methods, HsMM methods and RSS fingerprinting methods via a real experiment of a two-region emitter localization problem and Monte Carlo simulations using ray-tracing software.