Improved first responder localization for single-hop and multi-hop networks

Abstract

In this paper we present a novel algorithm for indoor localization that was developed for first-responder applications and which is based on range measurements between nodes. The common assumption that the range error is normally distributed is far from reality in indoor environments, where the range measurements exhibit a positive bias due to multipath. In our algorithm the range error distribution is estimated from acquired data and, by sampling this distribution, a probability distribution for the location of each mobile node is estimated. The distribution itself can then be taken as the solution for the node location or a specific node location can be extracted from the distribution in a variety of ways. We also extend this algorithm by constructing multihop constraint sets to perform multihop or cooperative localization, which increases the coverage in which the first responders can be tracked compared to a single hop network. The performance of the proposed algorithm is verified using data collected using a network of WASP (wireless ad-hoc system for positioning) nodes in an indoor office environment.