Adaptive bayesian beamforming with sidelobe constraint

The conventional Bayesian beamformer suffers substantial performance degradation, when the true direction-of-arrival is deterministic and is not included in the priori. In this letter, we propose a method with sidelobe constraint to improve the robustness of the Bayesian beamforming method. Support vector machine is used to obtain the weights. Numerical results show that the proposed beamformer can improve the Bayesian beamforming performance, and can output a relatively higher signal-to-noise-plus-interference ratio even when the desired direction-of-arrival is not included in the Bayesian priori region.