A Hierarchical Filter for Group Target Tracking Using Phased Array Radar

Multi-target tracking plays an important role in phased array radar(PAR) surveillance systems. In actual scenarios, targets often appear in the radar field of view(FoV) in formation. These targets may maneuver relative to each other and the formation changes such as splitting and merging during the motion, which poses a great challenge for PAR tracking. To solve the group target tracking problem for PAR system, this paper proposes a hierarchical filter, which uses two filters to control beam pointing and estimate the state of multiple targets within the group respectively. For multi-target filter, we propose a beam matching labeled multi-Bernoulli(BM-LMB) filter in order to apply multi-target tracker to the track-and-search(TAS) mode used in PAR tracking system. For the beam control filter, we introduce a state correction method to solve the easy loss of split targets in radar FoV. The multiple target state estimation is used as the ancestry information to provide accurate starting information for the new-born groups. Simulation shows that the algorithm proposed in this paper has a better performance for scenarios such as group target separation after introducing the correction method.