DOA Tracking Combining Compressive Sampling and Kalman Filtering for MIMO Radars Operating in Cluttered Environment

Taking into account the off-grid effect and clutter, a tailored compressive sampling (CS)-based algorithm is developed to estimate the Direction of Arrival (DOA) in radars that use Multiple-Input Multiple-Output (MIMO) technology and operate at sub-Nyquist rates. To enhance the performance of DOA estimation without resorting to denser sampling grids, which often entail expensive computational costs and robust sparse signal recovery issues, we incorporate the DOA estimates derived from the CS-based method into the state space of a Kalman filter. This integration eliminates the need for dense sampling grids, enhances computational efficiency, and improves robustness against clutter. Numerical experiments demonstrate that the proposed method consistently outperforms state-of-the-art techniques, achieving lower DOA error and superior receiver operating characteristic (ROC) performance.