A broadband application of memoryless narrowband GSC/NLMS adaptive beamformers

The adaptive performance of the normalized least mean-squares algorithm in the context of the generalized sidelobe canceller beamformer is considered. The implications of both the convergence behaviour and the misadjustment on various beamforming applications are discussed. In particular, an important case is identified for which there is near-instantaneous convergence. A misadjustment limit for which coherent post-processing is viable is also derived. Finally, a novel approach to coherent broadband beamforming is introduced and then tested via simulation