Distributed energy-efficient target tracking with binary sensor networks

Abstract

Target tracking is a typical and important cooperative sensing application of wireless sensor networks. We study it in its most basic form, assuming a binary sensing model in which each sensor returns only 1-bit information regarding target's presence or absence within its sensing range. A novel, real-time and distributed target tracking algorithm is introduced. The algorithm is energy efficient and fault tolerant. It estimates the target location, velocity, and trajectory in a distributed and asynchronous manner. The accuracy of the algorithm is analytically derived under an ideal binary sensing model and extensive simulations of ideal, imperfect, and faulty sensing models show that the algorithm achieves good performance. It outperforms other published algorithms by yielding highly accurate estimates of the target's location, velocity, and trajectory.