Atomic Decomposition-based Sparse Recovery for Space-Time Adaptive Processing

The requirement for a high number of training samples is a key factor that limits the application of space-time adaptive processing (STAP) technique in practice. In this paper, by exploiting the low-rank property of the clutter covariance matrix, we formulate an atomic norm minimization problem for the sparse reconstruction of the clutter-plus-noise covariance matrix. By defining the corresponding atomic norm in the continuous angle-Doppler domain, the proposed technique not only substantially reduces the number of required training samples, but also avoids the intrinsic basis mismatch problem that is encountered in conventional sparse reconstruction methods. Simulation results verify the effectiveness of the proposed STAP method and its superiority over existing techniques.