Tracking a Variable Number of Targets in Acoustic Sensor Networks by Particle Filter

A novel particle filter algorithm is proposed to track a variable number of targets in acoustic sensor networks, where the sensor data represent measurements of acoustic signals from one or more targets and background noise. The algorithm is based on the scheme of which assumed that at most one target may change between two time instants. The algorithm firstly estimate the number and identities of targets existing in the region, then estimate their states. The simulation results show that the new algorithm has good performance when tracking a variable number of targets.