Dimension-Reduced Space-Time Adaptive Clutter Suppression Algorithm Based on Lower-Rank Approximation to Weight Matrix in Airborne Radar

In this paper, we address the dimension-reduced space-time adaptive processing (STAP) techniques for ground clutter suppression in airborne radar from the viewpoint of approximation theory. The weights in the optimum STAP technique can be naturally expressible as the weight matrix. An efficient dimension-reduced space-time adaptive clutter suppression (STACS) algorithm based on lower-rank approximation to weight matrix is established, which finds a set of space-time separable filters to approximate the optimum STAP processor. By exploiting a lower-rank approximation to weight matrix, we make the quadratic cost function used in the classical optimum STAP processor be converted into a biquadratic cost function. To seek a minimum point of the biquadratic cost function, this paper develops an efficient multistage bi-iterative algorithm and the corresponding multistage dimension-reduced technique with a modular structure, where each stage finds an orthogonal component for approximating to the weight matrix. The effectiveness of the STACS algorithm is tested via several experiments.