Belief Consensus Algorithms for Distributed Target Tracking in Wireless Sensor Networks

In distributed target tracking in wireless sensor networks (WSN), agreement on the target state is usually achieved by the construction and maintenance of a communication path. Such an approach lack robustness to failures, and is not applicable to asynchronous networks. Recently, methods have been proposed that can solve these problems using consensus algorithms. However, these methods suffer from at least one of the following problems: i) they do not use fastest consensus methods, and ii) they cannot handle all parametric and nonparametric likelihood functions. In this paper, we propose a general framework for target tracking using distributed particle filtering (DPF) based on three asynchronous belief consensus (BC) algorithms: standard belief consensus (SBC), broadcast gossip (BG), and belief propagation (BP). Since DPF can be also solved (without consensus) by exchanging the observed data, we determine under which conditions BC-based methods are preferred. Finally, we perform extensive simulations to analyze the performance of these methods. Our main result is that DPF-BG and DPF-SBC provide the best performance in terms of root-mean square error, and that DPF-BP provides the best performance in terms of disagreement in the network. Manuscript received January 27, 2012. V. Savic and S. Zazo are with the Signal Processing Applications Group, Universidad Politecnica de Madrid, Madrid, Spain (Email: {vladimir,santiago}@gaps.ssr.upm.es). H. Wymeersch is with the Department of Signals and Systems, Chalmers University of Technology, Gothenburg, Sweden (Email: [email protected]). V. Savic is supported by the FPU fellowship from Spanish Ministry of Science and Innovation; This work is supported, in part, by the Swedish Research Council (VR), under grant no. 2010-5889; the European Research Council, under grant COOPNET No. 258418; the Spanish Ministry of Science and Innovation under the grants TEC2009-14219-C03-01 and TEC2010-21217C02-02-CR4HFDVL; program CONSOLIDER-INGENIO 2010 under the grant CSD2008-00010 COMONSENS; the European Commission under the grant FP7-ICT-2009-4-248894-WHERE-2. February 24, 2012 DRAFT ar X iv :1 20 2. 52 61 v1 [ cs .D C ] 2 3 Fe b 20 12