Multi-User MIMO Pre-Coding Technology Research For Wlan

With the rapid development of wireless communication technology and the spread of wireless terminals in recent years, as an important way of accessing the internet, wireless local area networks (WLAN) are widely used in enterprises, schools, families and other hotspots for its unique advantages. The explosive growth of wireless communication services and the number of users make people put higher demand on system capacity and user experience. Applying the Multi-user MIMO in WLAN can greatly enhance the system capacity and the number of accessible users, while it is bound to have serious multi-user interference when multiple users communicate with the access point(AP) in the same time and frequency resource. In the uplink AP can use multi-user detection technique to distinguish information from different users, while there is no collaboration between users in the down-link, therefore interference between multiple users only can be suppressed at the AP through using pre-coding technology.First of all, this thesis outlines the development of WLAN, the key physical layer technology as well as the multi-user data processing flow of the down-link. Then it analyzes the system model of downlink multi-user MIMO and the pre-coding mode in MIMO-OFDM system, then studies the commonly used linear pre-coding including the ZF pre-coding, MMSE pre-coding and BD pre-coding. Secondly, nonlinear pre-coding including the dirty paper coding and THP are researched, including the ZF-THP and MMSE-THP. It also combines with the Lattice Reduction technology to optimize the THP and improves the bit error rate performance, also makes corresponding simulation and analysis. Finally, this thesis highlights the pre-coding application in the WLAN. It analyzes the factors of resulting in imperfect channel state information including channel estimation error and frequency offset, then describes the channel state information feedback mechanisms and the method to reduce overhead of channel state information in the current WLAN and presents a multi-user parallel feedback scheme to improve the efficiency of feedback. Then it analyzes the influence of different quantization bits and the size of subcarrier group to feedback channel state information for pre-coding performance on pre-coding performance through simulations and presents a better way to select shared subcarrier channel state matrix for subcarrier grouping. At the end of this thesis, it proposes a idea of combining linear pre-coding and TH pre-coding according to their characteristics and present a scheme for changing multi-user frame structure and adaptive selection.