Multi-user Mimo Precoding And Power Allocation

Multiple Input and Multiple Output (MIMO) is one of the important wireless communications technology in the forth generation of the mobile communications. MIMO technology which can take full use of space resource will provide higher sum rate compared with the capacity of the system with single transmitting and receiving antenna. The capacity of the Multi-User MIMO system in which multiple users communication with the Base Station (BS) in the same channel is usually limited by the multiple users'co-channel interference. To compress the CCI, the transmitted signal is usually multiple with a precoding matrix to compress the CCI, then the user can receive the pure signal without interference.At first, the paper introduces the Dirty Paper Code (DPC),which provide high capacity as a same as that of the system without CCI. Since DPC is a non-linear precoding, it is difficult to realize in practical. Zero-Forcing (ZF) and Block Diagonalization (BD) precoding methods can cancel the CCI completely, but the capacity of the system with the above two precoding will decrease when the Signal to Noise Ratio of the system is very low. For that case, the paper provides a precoding methods, which balances the multiple users'interference and system noise, will improve the system capacity with low SNR.Besides designing the interference-compressing matrix, the propose method also guarantee each user (or stream)'s QoS, each user (or stream)'s minimum SINR requirement, by the power allocation. The power allocation will be finished in two steps, Firstly, allocate the power to guarantee the QoS with minimum sum transmission power. Secondly, allocate the remained transmission power to improve the capacity further. The simulation results show that the proposed method will achieve higher capacity than the existed linear precoding methods and the user (or stream)'s QoS will be met in the whole SNR region.In sum, the precoding and power allocation methods referred in this paper are discussed based on the condition that the ideal channel state information (CSI) is available. In the future, the paper will focus on the precoding methods with partly CSI or no CSI. On the other hand, the paper will extend the proposed precoding method in the single cell system to multiple cells system.