Adaptive channel regularization for multiuser multiantenna downlink

In the multiantenna downlink and under the availability of perfect channel state information (CSI) at base station (BS), regularized channel inversion (RCI) outperforms plain channel inversion (CI). Nevertheless, in the presence of imperfect CSI at BS, since RCI is more sensitive than CI to CSI imperfections, the performance of RCI becomes more degraded than that of CI. Therefore, in this paper, we propose an adaptive RCI technique which tries to compensate the degraded performance of standard RCI compared to CI when only imperfect CSI is available at BS. We do so by deriving an appropriate regularization parameter as a function of channel estimation error variance, which is possible to be known in advance. We also derive the output signal-to-interference-plus-noise ratio (SINR) of each user and show that the adaptive RCI is a generalized version of standard RCI such that by setting the error variance equal to zero, the former boils down to the latter. Simulation results certify that the proposed scheme results in improved bit error rates (BER) and higher achievable sum rates compared to standard RCI.