Performance evaluation of multi-user detection in CDMA using micro-genetic algorithm

Two main problems of multi-user communication systems are the multiple access interference (MAI) and the near-far effects. While the near-far effect problem can be approached by applying power control, the MAI problem requires individual receivers to identify the desired signal from interferences thus making multi-user detector a popular area of research. Genetic algorithms (GA) is a search and optimization technique that works by estimating multiple solutions in order to come with the best estimated solution. GA works iteratively over a population of solutions using crossover and mutation to simulate a potential solution. In this paper we examine the performance of a micro-genetic algorithm-based multi-user detector. Our simulation shows that /spl mu/GA achieves a performance close to an optimal detector.