Multi-Sensor Single Target Bearing-Only Tracking in Clutter

Multiple unattended ground sensors are deployed for surveillance, monitoring the movement of troops, military vehicles, and targeting. Usually, the probability of detection ( D P ) of an UGS is low and the false alarm density (FAD) is high. The particle filter (PF) and range-parametrized extended Kalman filter (RPEKF) have been used previously to produce improved results for the single sensor single target bearing-only tracking problem in the absence of clutter and with unity probability of detection. Although a great deal of work has been done for the single target single sensor bearing-only tracking problem, not much is known for the single target multi-sensor bearing-only problem with low D P and high FAD. We present algorithms and numerical results using particle filter (PF) and probabilistic data association (PDA) for the single target multi-sensor bearing-only tracking problem with high false alarms and low probability of detection. Our numerical results show that the algorithm estimates the target state in a robust manner in realistic harsh conditions when the probability of detection is low and the false alarm density is high.