The immune algorithm has been employed which receives efficient result.With the increasing need of limited frequencies, the second generation cellular network (Frequence Division Multiple Access, FDMA) is replaced by the third generation cellular network (Code Division Multiple Access, CDMA) due to its own inherent disadvantage. However, CDMA's inherent deficiencies, such as, multiple access inference (MAI) and near far effect (NFE), are fatal bottlenecks restricting its development. Multi-user detection technique can overcome or mitigate MAI and NFE so as to improve the receiving performance of the CDMA system. In this thesis, several kinds of intelligence multi-user detectors are introduced. Combining the endocrine mechanism and immune genetic algorithm, the immune-endocrine genetic algorithm (IEGA) is proposed. The convergence properties of the IEGA are proved that the IEGA converge to global optimum with probability 1. Then the IEGA was used in the intelligence multi-user detection. The experiment results demonstrate that this new algorithm has better performance than the traditional ones.
