Robust Adaptive Beamforming Based on Covariance Matrix Reconstruction Against Steering Vector Mismatch

Abstract

The conventional beamforming algorithms are sensitive to steering vector (SV) mismatch in practical terms, especially when the desired signal is present in training snapshots. It affects the received quality of the desired signal and the suppression of the interference, resulting in performance degradation. In this paper, a robust adaptive beamforming (RAB) algorithm based on interference-plus-noise covariance matrix (IPNCM) reconstruction against SV mismatch is proposed to address the problem. The proposed method performs a reconstruction-based IPNCM with respect to the Capon spectral estimator integrated over discrete angular sectors associated with the interferences. Subsequently, the eigenspace-based (ESB) projection method is employed to perform the correction of SV mismatch error by its projection on the signal plus interference subspace. The new method works on the refinement of the performance in high input signal-to-noise-ratio (SNR) and the desired signal SV mismatch as compared to most state-of-the-art RAB techniques, and it is able to provide similar performance close to the optimal value. These significant advantages are verified by numerical simulations.