| WCDMA is one of3G mobile standards. Interference from different users directly affected the overall system performance. How to eliminate the interference of multiuser has been an important research subject. We focus on the blind detection for multiuser interference in WCDMA because of the advantage that it has no needs for extra signalling transmission.In this paper, the blind multiuser detection algorithm for WCDMA system is studied. A blind multiuser detection algorithm which consists of3parts is designed——the average filter, MAI eliminate filter and MUI eliminate filter. At the beginning, the interference of WCDMA system is classify as two kinds:the MAI from users belong to the cell and the MUI from users from neighbour cells. Then, blind multiuser detection algorithm is reviewed.In order to design MAI eliminate filter, blind detection algorithms by using LMS,RLS and Kalman are studied. Furthermore, these algorithms are simulated and certified in both Rayleigh and AWGN channel environments. Because of the time-varying character, these blind detection algorithms become invalid on high-speed scenes. At the same time, there were performance reduced due to multipath effect.Therefore, these3blind detection algorithms must used after channel equalization filters.Then MUI eliminate filter is also designed. Since multiple users are independent and base station has no idea of any information from other sources, this blind detection can be seen as the blind source separation problem. The channel response matrix of MIMO-WCDMA is right the mixed matrix of blind source separation problem. ICA method is used to get the solution, the algorithm is simulated in AWGN channel.Finally, the introduced scheme which is made up of3filters is analysed, the limitation of which is also pointed. The distribution of WCDMA system speech signal will need further analysis.At the same time, to improve the blind multiuser detection algorithm requires a combination with the channel equalization technology. |
PDF Full Text Request