A Widely-Linear LMS Algorithm for Adaptive Beamformer

A novel adaptive Array structure using widely-linear weight is proposed in this paper. MMSE law is used to evaluate the performance of the adaptive array. According to the widely-linear structure, the widely-linear LMS algorithm (WL-LMS) is used here, which uses widely-linear model of the received signal rather than linear model estimating the desired signal. This structure provides a unified framework for beamforming in circular and non-circular sources, and can get much less MMSE between the beamforming output and the desired signal. This algorithm has been tested for beamforming using BPSK and narrow-band FM signal.