Significance of group delay spectrum in re-weighted sparse recovery algorithms for DOA estimation

Abstract

Sparse Recovery (SR) algorithms have been widely used for direction of arrival (DOA) estimation. At low values of signal to noise ratio (SNR) i.e. beyond -10 dB and with adequate number of sensors [1], their estimates are incorrect. The magnitude spectrum-based Re-weighted sparse recovery (RWSR) algorithms improve the robustness by re-weighting the sparse estimates. But their efficiency degrades significantly with a fewer number of sensors. The significance of phase spectrum in the form of Group delays for robust DOA estimation using RWSR algorithms for spatially contiguous sources is explored in this paper. An optimal re-weighting methodology based on simultaneously minimizing average root mean square error (ARMSE) and maximizing the probability of separation is proposed. The simulations are carried for Gaussian and Laplacian noise to demonstrate the superior performance of the proposed method with a few sensors at low values of SNR. © 2022 Elsevier Inc.