Estimation of multiple target location in multi-path wireless systems

Abstract

This paper addresses the problem of estimating the location of a target based on measurements of the signal run length from one transmitter, via the (passive) target, to multiple receivers. In particular we address the case that there are multiple targets that differ in their locations, but otherwise have the same characteristics, and thus cannot be distinguished by their "signatures" or other unique characteristics. Thus localization requires to correctly "pair" the time-of-arrivals (TOAs) of the reflections arriving at the different receivers. As a further challenge, in practical scenarios the number of the targets are unknown, and the wireless signal propagation can suffer from indirect paths (IPs). We present in this paper an algorithm that can localize the targets even under these difficult circumstances. Its steps include a clustering stage that gives the number and locations of targets, and an IP detection mechanism based on the likelihood of several TOA measurements being direct paths from the same target. The algorithm has much lower computational complexity than the exhaustive maximum likelihood estimation (EMLE), while our simulations demonstrate that it achieves almost the same performance as the EMLE.