Sidelobe suppression for adaptive beamforming with sparse constraint on beam pattern

Abstract

To suppress the sidelobe level of adaptive beamforming, sparse constraint on beam pattern is added to the conventional minimum variance distortionless response (MVDR) beamformer. The sparse constraint is an Lp-norm which can be explained as a diversity measure to evaluate the sparsity of the beam pattern. The proposed beamformer can not only reduce the sidelobe level, but also increase robustness against the mismatch problem caused by imprecise knowledge of the desired source.