Research On Blind Multiuser Detection Algorithms In DS-CDMA Communication Systems

CDMA has become one of the key techniques of 3G mobile communication system. However,multiple access interference(MAI) affects the system performance.Multiuser detection(MUD)is the way to restrain MAI.It is also good for restraining near-far effect and multipath interference. Blind multiuser detector has become an important study point as it has no needs of training sequence and the prior knowledge of interferential users.This thesis puts focus on blind MUD algorithms based on the current MUD technologies.Study the DS-CDMA system model of blind MUD. Make analysis and simulations the basic principle of minimum output energy(MOE) Least Mean Square(LMS) and Recursive Least Squares(RLS) two blind multiuser detection algorithm in synchronous DS-CDMA system under white Gaussian noise (AWGN) channel.On the base of deep research LMS algorithm, an improved variable step size LMS algorithm has been designed, simulation results show that performance of this algorithm is superior to LMS algorithm and it's calculation is almost same as LMS algorithm.The algorithm of RLS combined with decision-feedback variable tep-size MOE(DF-MOE) is proposed for dynamic environment.The result from simulation indicates that this algorithm has good suppression effect on dynamic strong MAI and near-far effect.Make analysis and simulatons on Signal Interference Rate in different channel environment of Kalman algorithm of blind MUD, and three blind adaptive multiuser detection algorithm are analyzed and compared,including blind adaptive LMS algorithm, RLS algorithm and Kalman algorithm. Based on subspace blind multiuser detection and Kalman algorithm, we get a improving type Kalman blind multiuser detection algorithm. The simulation of computer shows that is superior to Kalman algorithm in the convergence performance.