Sensor deployment for mobile object tracking in wireless sensor networks

Abstract

Mobile object tracking is a key application of wireless sensor network based surveillance systems. Sensor deployment is an important factor in tracking performance. In this paper, we consider the problem of optimal sensor deployment for mobile object tracking using the node sequences model without considering the sensing quality and the specific movement trace. We determine that there are three patterns that can reduce tracking error rate: regular tiling pattern, random pattern, and irregular pattern. We analyze the properties of these patterns and provide an upper bound of the tracking error rate for each pattern in the worst case scenario. We also provide theoretical analysis and simulation evaluations to demonstrate that the irregular pattern outperforms the other two patterns. © 2012 Springer-Verlag GmbH.