A Direction Constrained Eigenspace-Based Beamforming Algorithm

Abstract

The eigenspace-based (ESB) beamformer demonstrates much more robust capabilities than the minimum variance distortionless (MVDR) beamformer. However, when the pointing error falls in some certain positions near the mainlobe edge or the integer times of half of the mainlobe width, its performance will suffer from severely degradation. To avoid the certain positions, the paper proposed a modified algorithm by adding another distortionless constraint to the MVDR optimization problem in the direction different to but near the presumed direction. The new optimization can be solved as a linearly constrained minimum variance (LCMV) one. After solving it, the obtained optimum weight vector is projected onto the signal subspace. Several computer simulations are provided for illustrating the advantages of the proposed algorithm.