Distributed Target Tracking with Directional Binary Sensor Networks

Abstract

One of the most common and important applications of wireless sensor networks is target tracking. We study it in its most basic form, assuming the binary sensing model in which each sensor can return only information regarding target's presence or absence within its sensing range. However, unlike most of traditional approaches to binary sensing, we allow sensors to recognize not only target's range but also a sector within the circular range around it. Examples of such sensors include cameras, infrared sensors, ultrasonic sensors, etc. For simplicity, we assume that either a group of sensors are collocated in a single spot providing 360 degree coverage or a sensor has multiple antennas or camera providing such coverage. A novel, real-time and distributed target tracking algorithm with directional binary sensor networks is proposed. It is an extension of our previous work on omni-directional binary sensor networks. Using simulations, we demonstrate that this new algorithm achieves high performance and outperforms other algorithms by yielding accurate estimates of the target's location. In addition, we discuss the fundamental performance limits and improvement of the tracking performance resulting from providing direction range in addition to a distance range for the algorithm.