Research On Multi-User Detection Technology Of DS-CDMA System

Code division multiple access(CDMA)technology has many advantages,such as high bandwidth utilization,large system capacity,soft handover and good concealment security.It is widely used in modern mobile communications,navigation and positioning,radar and other fields.In direct spread code division multiple access(DS-CDMA)system,due to the limited resources of spread spectrum sequence,asynchronous transmission of information and random access of different users,the spread spectrum sequence of users can not always be completely orthogonal,and multiple access interference inevitably exists in the system,which results in the performance degradation of the system.The Optimal multi-user detection(OMD)technology has good detection performance,but with the increase of the number of users,the amount of computation increases exponentially,which is difficult to achieve in the actual system.This paper mainly studies the partial parallel interference cancellation(PPIC)algorithm based on linear decision function and the multi-user detection technology based on invasive weed optimization(IWO)algorithm.Aiming at the existing problems of the algorithm,some improvement measures are proposed.The main work is as follows:(1)The PPIC algorithm based on linear decision function is studied.An improved linear decision function is proposed and the parameter settings are analyzed.The simulation shows that the multi-stage PPIC based on improved linear decision function has good detection performance.(2)In order to apply IWO to multi-user detection problem,this paper transforms the weed diffusion value into the variation probability of a certain dimension of weed through the mapping function,and forms the binary IWO.Through this method,a multi-user detection model based on IWO is constructed.Aiming at the slow convergence rate of the normal diffusion standard deviation of the basic IWO,an improved standard deviation curve is designed,and a multi-user detection algorithm based on improved IWO(IIWO)is proposed.The effectiveness of the proposed algorithm in multi-user detection problem is verified by simulation.The performance of multi-user detection based on IIWO is close to the OMD,and it has better convergence speed.(3)Aiming at the problem that weed individuals do not share information and converge slowly in the invasive weed algorithm,this paper introduces particle swarm optimization(PSO)algorithm in the IIWO,and proposes a multi-user detection based on PSO-IIWO.Through simulation analysis,the multi-user detection based on the hybrid algorithm has better convergence performance than the multi-user detection based on the basic PSO and the multi-user detection based on IIWO.(4)Aiming at the problem that the multi-user detection is slow in convergence based on IIWO,the PPIC algorithm is embedded in the IIWO algorithm,and the multi-user detection based on IIWO algorithm embedded with PIC algorithm(IIWO-EPIC)is proposed.Through simulation analysis,the multi-user detection performance based on the hybrid algorithm is close to the optimal multi-user detection,and the convergence speed is faster than the multi-user detection based on the IIWO.