The Research On Power Control Strategy In DS-CDMA System By Genetic Algorithms

With the development of mobile communication, CDMA becomes a maintechnology for the third(3G) communication systems. There are many researches onCDMA system. However, efficient channel reuse is of critical importance in thedesign of the cellular wireless system of higher capacity. Co-channel interferencecaused by the frequency reuse in each cell is the single most restraining factor on thesystem capacity. Efficient power control schemes can eliminate Near-far effect andedge effect, reduce the interference form the same cellular, increase the systemcapacity, so power control has become a key technology in the third communicationsystems.In this paper, we propose an approach to solve the centralized power controlissue in CDMA system using genetic algorithms (GAs), to realize dynamicallocation of power. Nonlinear decreasing functions and FIFO stacks are introducedinto the simple genetic algorithm to speed up the convergence rate of power controlaccording to stochastic search and flexibility of GA. Then simulation onsmall-scale DS-CDMA cellular system and large-scale DS-CDMA cellular system.The result of the simulation show GA is effectively speed up the convergence rate,especially in large-scale DS-CDMA cellular system. The convergence rate ofalgorithms is an important factor to evaluation the algorithms of CDMA systempower control. We put the emphasis on the balance of services and convergence rate.Meanwhile, In large-scale DS-CDMA cellular system, because the mobile is notuniform distribution, which cause the mobile distribution is close to realisticenvironment.Though the study on centralized power control is based on theory, it can guidethe design the algorithms of distribution power control.