Sum-rate maximizing beamforming in multicell systems with limited feedback

Multicell cooperative transmit strategies typically require full channel state information (CSI) at the base stations. In this paper, a more practical limited feedback scenario is considered for a recently proposed multicell beamforming approach that approximately maximizes the sum-rates in a two-cell two-user network. In the limited feedback system presented, the desired and interfering CSI is quantized by the users and fed back to the base stations over a finite-bandwidth link. Each receiver is assumed to use two separate codebooks to quantize the desired and interfering channels. It is proposed to vary the codebook sizes adaptively as a function of the relative strength of the channels to minimize the loss in sum-rate. To this end, the mean loss in sum-rate due to random vector quantization is quantified. Closed-form expressions are derived to partition the available feedback bits adaptively between the desired and interfering channels to minimize the mean loss in sum-rate. Finally, simulations are used to verify the results presented in the paper.