Robust MDDR beamforming for sub-Gaussian signals in the presence of fast-moving interferences

The recently proposed minimum dispersion distortionless response (MDDR) beamformer is able to improve the reception of non-Gaussian signals over the minimum variance distortionless response (MVDR) beamformer. A scenario often encountered in practice is that a strong fast-moving interference crosses the observation region of an array, causing deterioration in the performance of the standard MDDR beamformer. To suppress such an interference, a continuous deep null sector over a predefined range of dynamic angle-of-arrival (AOA) is desirable. We develop a robust beamformer against fast-moving interferences based on the MD criterion. The proposed beamformer ensures that the output is optimized based on the MI) criterion and that the average power over the dynamic AOA is subject to be zero via a quadratic constraint. In order to reduce the computational load, the optimization problem with a quadratic constraint is relaxed to its counterpart with a linear constraint. This allows us to devise a fast algorithm, i.e., the Newton's method. Numerical results are provided to demonstrate that the developed beamformer can suppress the strong fast-moving interferences effectively.