ESB-DMR beamforming under random matrix theory judgment principle

Eigenspace-based beamformer and dominant mode rejection are robust against DOA mismatch and array model mismatch when SNR is proper. While the former algorithm suffers from performance break down at low SNR and at high SNR, the performance of the latter degrades. We propose a principle to judge the level of the SNR based on random matrix theory. The idea of random matrix theory is to identify the level of subspace swap phenomenon, which is closely related to input SNR. With the principle we are able to combine two methods together and both effectiveness and robustness of proposed principle are demonstrated by simulation results.