Adaptive wideband beamforming with robustness against presteering errors

We develop a wideband adaptive beamformer with an improved robustness against presteering errors. The proposed beamformer employs the presteered broadband array processor and models the uncertainty in the array manifold as norm-bounded phase errors at the outputs of the presteering delay filters. A state-space model is developed for this robust beamforming problem, and an extended Kalman filter (EKF) is used to estimate the beamformer weight vector adaptively with a low computational complexity.