Recursive Bayesian beamforming with uncertain projected steering vector and strong interferences

Abstract

A recursive Bayesian beamforming is proposed for the steering vector uncertainty and strong interferences. Signal and noise powers are unknown, and beamforming weight is modeled as a complex Gaussian vector that characterizes the level of projected steering vector uncertainty. By applying the Bayesian model, a recursive algorithm is developed to estimate beamforming weight. Numerical simulations of linear and planar arrays demonstrate the effectiveness and robustness of the proposed beamforming algorithm. After convergence, the proposed algorithm exhibits a performance similar to that of the optimal MaxSINR beamformer.