Sparse Bayesian Learning-Based Space-Time Adaptive Processing With Off-Grid Self-Calibration for Airborne Radar

Space-time adaptive processing (STAP) for airborne radar has recently been enriched owing to the development of methods based on sparse recovery techniques. These methods have shown advantages over the conventional ones. However, there are still difficulties in practical situations, for example, when many clutter components are not located on the discretized sampling grids of the dictionary, which will result in significant performance loss. To deal with such off-grid problem, this paper proposes a sparse Bayesian learning-based STAP (SBL-STAP) with an off-grid self-calibration method, which can effectively mitigate the off-grid effect. In the proposed method, the clutter plus noise covariance matrix is estimated via SBL. Meanwhile, we construct a small-scale complementary dictionary with an adaptive approach to calibrate the uniformly discretized dictionary. In each iteration of the SBL, the atoms of the complementary dictionary renew themselves by an approach based on weighted least squares. In this way, the atoms of complementary dictionary can converge to the clutter ridge adaptively when off-grid occurs. The simulation results show that the clutter ridge spreading caused by off-grid can be mitigated effectively, and the output signal-to-clutter-plus-noise ratio of the STAP is significantly improved. The benefits come at the cost of negligible increase of computational complexity.