Robust Widely Linear Reduced-Rank Adaptive Beamforming Based on Joint Iterative Optimization

Abstract

Based on joint iterative optimization, a robust widely linear (WL) reduced-rank minimum variance distortionless response (MVDR) beamforming algorithm is proposed for noncircular signals. To handle the issue of the mismatch between the presumed array steering vector and actual array steering vector, a diagonal loading technique is devised. As the proposed beamformer takes advantage of the noncircular properties of source signals, it performs better than the reduced-rank robust joint iterative optimization (JIO) and the WL reduced-rank JIO adaptive beamformers. Simulation results shows that the proposed beamformer provides faster convergence speed, larger output signal-to-interference-plus-noise gain and less mean square errors than the existing approaches.