International Journal of Ad Hoc and Ubiquitous Computing
| Hi-index | 0.00 |
Over the last few years, target tracking in wireless sensor networks has become a topic of particular interest. This paper presents a tracking system intended for deployment in distributed Wireless Sensor Networks. The approach is inspired from the concept of deploying sensor nodes in an ad-hoc manner and based on the aggregate amount of information they provide, perform tracking of mobile objects. Static sensor nodes, with known locations, act as anchor nodes providing range-only measurements at each time step. The proposed system is modeled using a nonlinear state-space model which reflects a real world tracking scenario. The batch of range measurements made available at each sampling step, is used to estimate the target's desired kinematic properties. In order to infer the state of the target at each time step, a Particle Filter algorithm has been designed to approximate the required posterior distribution of the state vector. The system's operation was simulated and execution examples demonstrate the algorithm's accuracy as well as the ability to effectively cope with manoeuvring targets.