Adaptive Widely Linear Constrained Constant Modulus Reduced-Rank Beamforming

Abstract

We propose a blind adaptive reduced-rank widely linear beamforming algorithm using the generalized sidelobe canceller structure for interference suppression. A structured Krylov-subspace-based approach is devised to construct the dimensionality reducing transformation matrix, and a recursive least-squares algorithm is developed according to the widely linear constrained constant modulus criterion to update the reduced-rank filter. We analyze the convergence and complexity of the proposed algorithm and validate its performance gains through simulations.