Tracking manoeuvring mobile nodes in wireless sensor networks with range-only measurements

Abstract

This paper addresses the problem of tracking mobile nodes in Wireless Sensor Networks (WSN). Our approach is based on a tracking with range-only measurements scheme, intended for implementation in WSN. The developed tracking system is designed for continuous estimation of the target's trajectory and two-axes velocity. A network of "anchor" wireless nodes is considered to be deployed at a specific region of interest. The anchor nodes collect data that correspond to the range between the anchors and the target. A multiple-modal Particle Filter (PF) algorithm is designed to process the ranging observations and produce an estimation of the target's kinematic variables in real-time. The reason for employing multiple models to represent the target's motion pattern, stems from the need to effectively track manoeuvring targets. Manoeuvring targets require more complex modeling which adequately represents the sudden changes of the position and velocity vectors. Simulations are provided to assess the performance of the proposed framework, considering real-world conditions. The proposed multiple model approach is evaluated on manoeuvring targets with constant as well as variable turning rate. Finally, this investigation reveals that the system's accuracy depends upon two important system parameters, the sampling period and the number of generated particles in the PF algorithm. The effect of these parameters is analyzed and amendments are proposed.