Transmit Beamforming and Detection Design for Uplink Multiuser MIMO Systems

We consider the transceiver design for uplink multiuser multi-input multi-output (MIMO) communications, where multiple transmitters adopt beamforming to simultaneously access the linear spatial equalizer at the receivers. In this paper, we investigated the crucial, yet not well discussed issue to jointly design beamforming weights for all transmitters, subject to total transmission power constraints, under the assumption that all users' channel state information (CSI) is available at each of the users. We proposed a class of orthogonal joint beamforming design schemes with matched filter receiver, and an eigen-beamforming (Eig-BF) scheme with ordered successive interference cancellation minimum mean square error receiver (OSIC-MMSE). By comparing these schemes with the optimal joint transceiver design with optimality in terms of overall mean square error (MSE), we show that Eig-BF OSIC-MMSE scheme achieves near-optimum performance at much lower complexity and with much simpler transmitter design, since it only requires each transmitter to know its own CSI and thus can be implemented in a decentralized way. This indicates that marginal performance improvement can be benefited from making all transmitters share their CSI. The orthogonal beamforming schemes, on the other hand, accomplish the simplest receiver structure at the cost of more complicated cooperative transmitters, and thus they may find applications in such systems as satellite communications or ad-hoc networks, where it is better to distribute computational loads among multiple transmitters than to impose it on a single receiver.