Finite Rate Channel Feedback Design for Multiuser MIMO Downlink Beamforming Systems

An efficient finite rate channel feedback design for multiuser multi-input multi-output (MIMO) downlink beamforming systems with individual signal-to-interference-plus-noise ratio (SINR) constraints is proposed. We find that part of the decomposed channel matrix is irrelevant to the performance, and can be discarded to reduce the feedback load. The remaining information is further decomposed into two parts, and we find that unevenly quantizing these two parts is better than traditionally quantizing the whole matrix evenly. We then show that linearly scaling the number of quantization bits for only one of these two parts is sufficient to maintain a bounded system performance loss. Thus quantization bit allocation between these two parts can be easily done for different signal-to-noise ratios (SNR). Simulation results verify our analysis and show that the proposed design outperforms the existing scheme.