Robust adaptive beamformer based on minimum dispersion with a broad null by linear programming

The standard minimum dispersion distortionless response (MDDR) beamformer can provide a performance improvement for non-Gaussian signals compared to the minimum variance distortionless response (MVDR) beamformer. However, the variety of the direction of the interference may have a great influence on the performance of the MDDR beamformer. To solve the problem, a robust beamformer with a broad null by linear programming is proposed for sub-Gaussian signals. Unlike the beamformers based on the minimum variance criterion, the proposed beamformer minimizes the
$\ell_{p}$
-norm of the output and meanwhile constrains not only the target response to be unity, but also the average power of the interference's dynamic angular sector to zero. The resulting optimization problem is cast as a linear programming and hence can be solved efficiently. Simulation results indicate that the proposed method offers a significant performance improvement if the direction of the interference varies with time.