Instantaneous and average rate maximization in MIMO multiple-access channels (MAC) with linear processing

Linear precoding (beamforming) techniques provide simple processing methods with good performance at reduced signaling complexity compared with nonlinear precoding schemes. The beamforming technique that maximizes the achievable sum-rate in the multi-user, multiple-input multiple-output (MIMO) broadcast channel (BC) has the distinct form of a regularized channel inversion (RCI). This current work adapts the method used for BC sum-rate maximization to maximize the sum-rate in the MIMO multiple-access channel (MAC) under fixed, per-user power constraints. This beamformer is based on accurate channel state information (CSI) resulting in performance degradation when only erroneous CSI is available at the receiver. Consequently, we develop an additional beamformer based on channel distribution information at the transmitter and/or receiver that attempts to maximize the average sum-rate, allowing for a stability/throughput tradeoff in the MAC.