Direct Data Domain STAP Based on Atomic Norm Minimization

This paper presents a new direct data domain (D3) STAP method established in the continuous domain for an airborne radar system, which can effectively suppress the clutter within the cell under test (CUT) data only. The proposed algorithm applies the oblique projection to elimate the potential target component from received data in advance. Then we reconstruct the CCM via solving the atomic norm minimization problem, without off-grid problem or loss of system degree of freedom (DOF). In addition, the reconstruction is relied on the positive semidefinite (PSD), low rank properties of the clutter covariance matrix (CCM), and the prior structure of block-Toeplitz. In order to further improve accuracy of the CCM estimation, we utilize the orthogonal subspace of the CCM to limit the range of the clutter subspace. Without training data available, the novel solution is able to deal with the non-stationary clutter scenario more excellently. In the end of paper, the simulations demonstrate that the proposed algorithm can estimate the CCM more accurately and acquire better performance of clutter suppression than other typical D3-STAP methods.