Multiple-Model Approach to Over-the-Horizon Radar Tracking Based on Target Perceivability

For over-the-horizon radar (OTHR)-based target tracking, the reflecting height of the ionosphere which reflects radar signals is important. Existing methods assume that this height is exactly known a priori. In practice, however, we can only determine its range, not its value. To circumvent this problem, we propose to use a multiple-model approach in which each model corresponds to a specific height for OTHR tracking. Further, existing methods do not provide a systematic and effective solution to track confirmation and termination, and mean-square-error (MSE) matrices of state update calculated by these methods are in error. To overcome these drawbacks and estimate the target state, an autonomous multiple-model target perceivability (TP)-based multipath probabilistic data association (TP-MPDA) tracker is proposed, where TP probabilities are derived for track confirmation, termination, and refinement, and MSE matrices of state update are corrected. Seeing that a small fixed set of models does not cover a large height range well, an expected-mode augmentation TP-MPDA tracker is proposed, which augments a basic model set by an online adaptive expected height. Since the model is nonlinear in the height, we also present a best model augmentation TP-MPDA tracker, where a basic model set is augmented by the best model that minimizes a Kullback-Leiber divergence. Simulation results demonstrate that the proposed algorithms are effective and have better performance than the existing MPDA algorithm.