Based On Particle Swarm Optimization And Its Hybrid Algorithm Of Multi-user Detection

The multi-user detection (MUD) is the key technique for Code-DivisionMultiple-Access (CDMA) mobile communications system. In the actual CDMAcommunication system, multiple access interference (MAI) is induced by multi-pathpropagation which destroys the orthogonality among the spread signals. The MAIcaused by the individual user is very small. However, with the increase of the users orthe enhancement of the signal power, MAI is becoming the main interference in thesystem. Consequently MAI has been becoming the fatal bottleneck of the furtherdevelopment of CDMA system.While there are a number of users in the system, the conventional detection (CD)demodulates each user's signal as if there is only one in it. So the method resistingMAI is not satisfied. The performance of MUD is much better, which makes systemmore efficiently by using all the user's information to detect the single user. Thismethod can resolve the near-far problem, and make the system need lower request forthe power control. So we can utilize the uplink frequency spectrum resourcesefficiently and enlarge the system capacity saliently.In 1995, Eberhart and Kennedy reported that the particle swarm optimization(PSO) algorithm was a parallel optimization. Since the PSO is proposed, more andmore scholars in the field of evolutionary computation have paid attention to it,because it has profound intelligence background and is easy to implement.In this thesis, The PSO strategy was applied to the search process of the MUDbased on the fact that MUD can be viewed as a combinatorial optimization problem.Comparing the corresponding relation between the fitness function of PSO and theobject function of optimal detector (OMD), one multi-user detector based on PSOwas proposed. Considering that PSO is easy to trap in local optima and can't find thebest efficiently, we proposed and studied two MUD techniques on the basis ofhybrid optimization strategy. One is the multi-user detector based on particle swarmoptimization and multistage detection (PSO-MSD), the other is the multi-userdetector based on particle swarm optimization and tabu search (PSO-TS). Numericalsimulation, analysis and discussion are made among the above detectors OMD, CD,...