Non-Cooperative Target Localisation Using Rank Based EDM Approach

摘要

An analysis of the effect of the number of sensors on the non-cooperative target node localisation is presented. This work examines the target localisation using a centralised range based approach. In doing this, the work leverages an algorithm based on a class of matrix structure called Euclidean distance matrices (EDMs) for the specific purpose of improving localisation performance when the fusion centre (FC) cannot receive certain sensors' information due to fading or shadowing, etc. While this interesting approach to the problem of localisation has been found to be successful, it is also shown at high delay values the proposed alternating rank-based EDM algorithm outperforms the conventional linear least squares (LLS) based algorithm for the minimum number of sensors. The localisation error decreases when the conditioning of EDM is better, i.e., when the sensors are further apart from each other and closer to the target.