A Study On Blind Muti-User Detection Algorithms In Direct-Spread Code-Division Muti-Access Systems

In code-division multiple-access(CDMA) systems, multi-access interference(MAI) is the predominant factor that determines the system performance as well as capacity. However, MAI is not inevitable. It can be alleviated or even eliminated by using multi-user detection which improves the system performance and enhance its capacity. Recently, blind multi-user detection has been the research focuses. Based on the existing MUD technologies, in this thesis, blind MUD algorithms are studied in theories and through simulations. The main contributions of this thesis are as follows:(1) The conventional Minimum Output Energy(MOE) algorithm has the global convergence but its stable excess mean square error is high and it can't converge to the Minimum Mean Square Error(MMSE). In this thesis an equivalent decision feedback MOE(DF-MOE) algorithm variable step-size is presented and analyzed. Computer simulation results show that the convergence performance of the new algorithm are better than those of MOE detector and Sato detector. This detector keeps the global convergence of the MOE detector and has an ideal stable state output signal-interference rate. It can overcome the MAI effectively.(2) To enhance the tracking performance of blind averaged stochastic gradient(BAG) algorithm, two BAG algorithms is presented based on the MOE criterion. One is the tracking BAG(TBAG). Based on the TBAG algorithm, an improved algorithm with the DF-MOE variable step size is presented. Computer simulations show that the two algorithms have good convergence and high stable output signal-interference rate(SIR) and tracking capability.