A Subspace Based Blind Multi-user Detection In DS-CDMA System

In CDMA mobile communication system, because of the multi-user's randomly connection,the spread spectrum codes they used are normally not quadrature rigidly. The nonzero crosscorrelation coefficient between codes will arouse the interference between the users, which iscalled Multiple Access Interference (MAI). The existence of MAI can severely limit the capacityof the system, and also make the strong signal submerge the weak signal. These days,DS-CDMA communication system mainly uses power control technique to restrict MAI, but thistechnique can only relax MAI, it can't solve MAI problem. In order to restrict MAI effectively,multi-user detection has become a hot research field in the communication system. The mainidea behind it is to make joint detection on the received signals by making full use of theinformation of all the users' signals, which eliminates the multiple access interference (MAI) andrelaxes the 'near-far' effect, and consequently increases the capacity of the system. This essaypresents a research of two parts.Firstly, the parameter and the performance of the conventional single-user detector(CD),theoptimal multiuser detector(OMUD),the decorrelating detector(DD) and MMSE blind multiuserdetectors have been analyzed in this paper.Secondly, a blind adaptive Kalman multi-user detection algorithm based on the improvedsubspace tracking has been proposed on this paper. Theory analysis of this algorithm and theperformance analysis of Matlab simulation are also given in this paper. Simulated results showthat it can follow the changing signal space rapidly and achieve better tracking SINRperformance while the channel changes suddenly.