Parameter-searched OMP method for eliminating basis mismatch in space-time spectrum estimation

The basis mismatch in the sparse recovery based space-time adaptive processing (SR-STAP) can severely affect the reconstruction of the space-time spectrum. In this paper, we propose a parameter-searched orthogonal matching pursuit (PSOMP) algorithm to eliminate the effect of the basis mismatch in SR-STAP. In the PSOMP, the optimal parameters carried in the atom selected by the OMP are searched so that this atom can match with the true steering vector of the corresponding scatter in the data. The feasibility of the key step of the PSOMP is theoretically analyzed. Finally, simulation results demonstrate that the PSOMP has a high performance in eliminating the effect of the basis mismatch in SR-STAP.