Precoding For Multi-user MIMO Channel

In multiple input multiple output(MIMO) systems,the precoding schemes at the transmitter,which uses the channel state information (CSI),are proposed to improve the spectrum efficiency and the power effiency. While a major problem in multi-user MIMO systems is the cancellation of the interference arising among different spatially multiplexed signals directed to individual users. For the downlink,since the mobiles are decentralized, the users are not able to coordinate with each other and can not separate the signals transmitted by the base station using array processing,multi-user detection (MUD), or other methods. Precoding algorithms for multi-user MIMO systems,which uses the channel state information (CSI) to process the transmit signal, have to be used to pre-eliminate the spatial interference among different mobiles. Then reduce the complexity and cost of the receivers and have great value.Linear precoding scheme was first introduced in this paper. An linear precoding scheme, which can completely eliminate the interference the signal,was introduced. It is a very good solution for low noise and high power situations. However ,it has been shown that it does not result in the linear capacity growth with min( nt , nr ) that should be achievable in the multi-user channel. Allowing a limited amount of interference at each receiver which based on MMSE criterior was secondly introduced. It dose achieve linear growth in throughput with min( nt , nr ),but at a rate that is somewhat slower than that for capacity.The concept of"dirty paper coding"and nonlinear precoding technology were also introduced in this paper. The nonlinear precoding technology are primarily Tomlinson-Harashima precoding based on zero-forcing and MMSE criterior and vector precoding. Although it is much more complicated to implement compared with linear precoding , it can achieve the upper boundary of a Multiantenna Gaussian Broadcast Channel. Besides,the performance of the system can be improved by ordering the rows of the channel matrix H .And"best-first"ordering ,which is proved to be the optimal ordering method ,is used in this paper.