SSP'05 -- Proceedings

	IEEE/SP 13th workshop on Statistical Signal Processing July, 17-20, 2005 - Bordeaux - France

Information regarding the paper

Title

Distributed Tracking in Multi-hop Networks with Communication Delays and Packet Losses

Author(s)

Murat Alanyali	Boston University
Venkatesh Saligrama	Boston University

Get the paper in PDF format

To obtain Acrobat Reader (version 5 minimum required) necessary to his read.

Abstract

Distributed tracking of a dynamical system via networked sensors is considered. A Gaussian signal estimation framework is adopted. The networked sensors communicate with each other by means of a multi-hop protocol. The multi-hop protocol leads to informational delays. MMSE-optimal solution is Kalman filtering when measurements are available centrally, but it is infeasible in the presence of communication delays within the network. The associated loss of performance is quantified and local message passing algorithms are proposed to allow each sensor to obtain a best feasible estimate. The proposed techniques differ from related techniques proposed in two important aspects: a) there is no designated leader/fusion node and each sensor attempts to optimally track the system trajectory based on its local observations and time-dependent information available from other sensors in the network; b) the message computation at each sensor is structurally identical, where the computed message from sensor, J, for a neighboring sensor, K, is the innovation from all sensors that are closer to J than K (in terms of communication delay). In turn, these results imply that the sensor network can be queried at any time and any place to obtain optimal estimates for the state of the dynamical system.