Particle Swarm Optimization Algorithm Is Applied Research In Multi-user Detection

The particle swarm optimization (PSO) algorithm was invented by Eberhart and Kennedy in 1995 as a parallel optimization. It comes of the behavior of birds searching food. Since the PSO is proposed, because of its advantage in easy implementing, visualizing and intuition, good at convergence, thus it is widely researched. In the international conference of evolutionary computation, PSO is regarded as important as other evolutionary algorithm. The researches on PSO mainly focus on two aspects. One is the applying research. The other is the studying of itself. Through ten years'developing, there are a lot of productions widly used in many fields.The Code-Division Mutiple-Access (CDMA) is always disturbed by sorts of interferences. In tte actual CDMA communication system, multiple access interence (MAI) is resulted from multi-path propagation which destroys the orthogonality among the spread signals. Multi-uesr detection (MUD) refers to signal detection in multiple access communications. This detection has many good properties, such as decreasing multiple access interference, canceling out or moderating near-far resistance, able to decrease the requirement for output control, compensating the effects due to inoptimal orthogonality, improving system performance and increasing system capacity. MUD is one of the methods that can overcome the MAI in CDMA with most potential and complete and becomes one of the critical techniques in the third generation mobile communication.This paper firstly overviews the traditional PSO algorithm, and introduces the stretching technique to avoid "premature" in PSO and to improve the algorithm's ablity to search for a global optimum. The stretching technique is applied to the optimization process in MUD. By viewing the objective function of MUD as the algorithm's fitness function, the multi-uesr detector based on the stretching PSO (SPSO) is proposed. Secondly, in order to improve the local searching capability, the modified PSO (MPSO) which jumps according to probability is introduced, and the MUD based on MPSO is proposed. Finally, for the purpose of the further improvement of the MUD performace, a third detector based on the mixed PSO (XPSO) is proposed. Through simulating the...