A polarization information aided probabilistic data association for target tracking in polarimetric radar system

In conventional radar tracking, kinematic measurements such as range, bearing and Doppler are usually used in target tracking methods. It has been demonstrated that the tracking performance can be further improved by utilizing more information. With regard to polarimetric radar system where polarization information (PI) is available, PI thus should also be exploited to ensure robustness of tracking performance. This paper considers the target tracking of polarimetric radar by incorporating PI into the probabilistic data association filter (PDAF). Firstly, a generalized structure of PI aided PDAF (PDAFPI) method is given. It is suitable for different radar models and different data association algorithms. Secondly, for the polarimetric monostatic radar, two versions of PDAFPI method are proposed for a single-target tracking. Specifically, the first version is the PDAFPI algorithm in optimum case (O-PDAFPI) where the polarimetric characteristics of target and noise covariances are known as a prior. The later one is a fully adaptive PDAFPI algorithm suitable for the suboptimum cases (S-PDAFPI) where the polarimetric parameters are unknown. Simulation results show that the proposed algorithms can effectively improve the tracking performance and track more reliably both non-maneuvering and maneuvering targets.