Robust transmit beamforming based on imperfect channel feedback

Transmit beamforming is a powerful approach to improve the performance and increase the throughput of wireless communication systems employing multiple antennas at the transmitter. However, most existing transmit beamforming techniques require nearly perfect knowledge of the channel at the transmitter, which is difficult to obtain in practice. Transmitter designs based on statistical models were introduced to address the imperfect channel state information (CSI) problem. Such model-based designs are susceptible to performance loss as a result of mismodeling. In this paper, we propose a more robust framework in which we do not make any statistical assumptions about the CSI mismatch or the channel. We design our robust transmit beamformer in order to have the best performance under the worst-case CSI mismatch. This beamformer turns out to be an eigen-beamformer along the eigenvectors of the spatial channel correlation matrix. Power loading across the beams is determined in a closed form by a spatial water-filling-type strategy.