RF-Aware Widely-Linear MMSE Beamforming

Beamforming offers good spatial resolution for the wireless communication systems and thus enables sophisticated transmission and interference suppression scenarios. In this paper, data-aided receiver beamforming is studied under practical limitations in the associated radio frequency (RF) circuits, namely the amplitude and phase imbalance between the inphase and quadrature-phase (I/Q) rails. Closed-form analysis of the available beamforming capabilities is provided, showing that the beamformer performance degrades heavily due to the I/Q imbalance. Motivated by the widely-linear (WL) structure of the I/Q imbalance, we then propose deploying data-aided WL minimum mean squared error (MMSE) beamforming solution for the beamforming task. Simulation results show that the proposed WL beamforming solution is able to efficiently suppress the unwanted I/Q imbalance effects and thus provides good overall performance, even under severe imperfections in the RF circuits.