IMM Filter based Sensor Scheduling for Maneuvering Target Tracking in Wireless Sensor Networks

This paper discusses the problem of maneuvering target tracking in a distributed wireless sensor network (WSN). In many applications, such as the human motion tracking or military/civilian surveillance, the target often moves with high maneuvering, and appeal to more advanced tracking approaches. The limited energy constraint in WSN complicates the problem further. In this paper, we present an interactive multiple model (IMM) filter based collaborative maneuvering target tracking framework that incorporates a novel energy-efficient sensor scheduling scheme in a distributed WSN using low cost range wireless sensor nodes. The proposed algorithm applies the IMM filter to estimate and predict the target's dynamic state and select the tasking sensor node and sampling interval for each time step based on both of the tracking accuracy and the energy cost. Simulation results show that the proposed approach outperforms the popular extended Kalman filter (EKF) based tracking scheme for maneuvering target in terms of tracking accuracy and energy efficiency.