Research On Precoding Techniques For Downlink In MIMO Systems

Multi-Input-Multi-Output (MIMO) system is a key technology of future broadband wireless communications, and is an important mean to make full use of space resources to improve spectrum efficiency. One of the main advantages MIMO systems is the essence improvement of channel capacity, which directly translates into high data throughput. Another advantage is remarkably enhances the reliability of data transmission, namely lower BER. These benefits don't cost of increasing the available bandwidth or higher transmit power. In downlink channel, MIMO precoding concept is proposed to reduce the inter-symbol interferences and inter-user interferences, reduce the complexity of the receiver. As countries around the world to research and perfection the technique continuously, precoding technology must be the key technologies that will be used in MIMO communications systems.This paper first discusses the background, the significance, the development of history and the present situation of researched precoding technology; then intensive research on the basic principle, system model, system capacity, spatial diversity and spatial multiplex of MIMO system. In the Matlab environment, system capacity is carried on the detailed simulation and analysis, and lays a foundation for studied the encoding algorithm.In single user mode, the paper first introduces the system model of precoding, then research TxZF and TxMMSE precoding schemes, and the solution process of TxMMSE precoding is given. Then SVD and GMD channel decomposition technique is studied. Then the equivalent channel is constructed based on MMSE criterion,and precoding matrix and equilibrium matrix are obtained by the GMD of the equivalent channel. In multi-user mode, the paper introduces the BD and BD-GMD algorithm. According to BD-GMD decomposition, a sub-channel sorting algorithms is introduced, and combines the sorting algorithms with BD-GMD method. Then new precoding matrix and equilibrium matrix are obtained based on minimizing transmitted power criterion. In the Matlab environment, the studied algorithms are simulated and analyzed in order to specific understand the advantages and disadvantages of each algorithms. At the end of the article, we sum up the main contents and works to MIMO precoding,and also analyzes the future work direction in the field of MIMO precoding .