Novel BD MU-MIMO Pre-Coding Methods to Suppress Noise and Balance Receive Antennas

This paper focuses on transmit pre-coding for multiuser multi-input multi-output (MU-MIMO) downlink channels. Multi-user interference (MUI) and noise are main factors that affect the performance of MU-MIMO system. Besides, the system performance will degrade as the receive power on different antennas gets unbalanced. The Block-diagonalization (BD) algorithm eliminates the MUI completely. However, the number of transmit and receive antennas are strictly limited, and the mitigation of noise is not considered. In this paper, two novel methods are proposed to solve these problems. One is antenna balancing in which we evaluate the receive power of each antenna and discard the unbalanced ones. This method could also relax the constraint on the number of antennas and enable more data streams transmission. The other one is noise suppression in which we design the pre-coding matrices using equivalent channels with noise compensation. Simulation results reveal that the noise suppression method can achieve better performance than BD, especially at lower SNR scale and the antenna balancing method can outperform BD in a wider SNR range. The methods provide significant gain in capacity without increasing computational complexity and greater performance improvement can be acquired when the two proposed methods are combined.