Adaptive multi-target tracking in heterogeneous wireless sensor networks

Abstract

Energy efficient multiple-target tracking is an important application of Wireless Sensor Networks (WSNs). Most prior studies consider tracking multiple tar- gets as an extension of executing a single target tracking algorithm multiple times, and use a single parameter for energy efficiency. We consider various factors such as mul- tiple targets tracked by the sensor, remaining energy of the sensor and relative location of the sensor with respect to a target's motion, in order to decide the tracking state of a sensor in a distributed environment. Further, we explore and identify the effective combination of these parameters to optimize energy usage, depending on specific network conditions. We then propose the Adaptive Multi-Target Tracking (AMTT) algorithm that can recognize the network condition based on local information without centralized coordination, and uses effective parameters to achieve en- ergy efficiency.