Precoded dirty-paper coding with signal-to-interference plus noise ratio-based ordering for multiuser multiple-input multiple-output communication

Abstract

The authors propose a joint dirty-paper coding (DPC) and beamforming scheme for use in multiuser multiple-input multiple-output (MIMO) systems. Unlike conventional maximum likelihood (ML)-based DPC schemes, this joint DPC and beamforming precoding scheme is intended as a suboptimal strategy in that it cancels only causal interference, thus offering a practical implementation option. In addition, a signal-to-interference plus noise ratio (SINR)-based ordering strategy that can be readily applied to conventional DPC schemes is also proposed as a way of improving the performance of the joint precoding scheme. The proposed schemes markedly improve classical spatial division multiple access (SDMA), and achieve the same data rates as spatial multiplexing (SM) for all users, but with significantly superior performance/diversity gain. When the number of active users per sector is much greater than the number of transmit antennas, the proposed schemes are able to achieve enough multiuser diversity to asymptotically approach the optimal DPC capacity. In addition, unlike BLAST, the receivers do not need to know each other's vector channels. Simulation results confirm that the proposed schemes provide significant gain over conventional zero-forcing SDMA in terms of average sector throughput for downlink multi-cell multiuser systems and symbol error rate as well. Finally, the proposed interference cancellation strategies at the transmitter can be expandable to MIMO systems with any number of multiple antennas.