Improved Density Assisted Particle Filtering for Target Tracking in an Asynchronous Wireless Sensor Netwok

Target tracking in a wireless sensor network (WSN) has become a relatively standard problem. In an asynchronous WSN, combined estimation of the target state and static clock offsets is necessary as the local clocks of the sensors are misaligned and the corresponding offsets are unknown. In this paper, a new tracking algorithm based on density assisted particle filtering (DAPF) and a sliding window is proposed. The sliding window calculates the reliability of the estimated sensor offsets and decides when to stop estimating them. Simulation results indicate that the improved DAPF algorithm we proposed can achieve high accuracy as standard DAPF but effectively saves the computing cost.