Study Of A Decorrelating Multiuser Detection Based On Genetic Annealing Algorithm

Code division multiple access(CDMA) communications systems being possessed of the merits of great capacity and soft hand-off and so on must have taken a significant role in future wireless communications systems. Because PN sequences used in communications systems are been vertical completely , when numbers of user are more, there are multiple access interference(MAI) in CDMA and serving quality in communications systems is reduced. If interference are avoided and capacity of systems is increased, multi-user detection technology must be used in CDMA.This paper introduces the application of the spread spectrum technology in communications. Aiming characteristic in CDMA communication systems, after we discuss some multi-user detection technology at present, and want to find a effective multi-user detection method with a small bit error rate in CDMA communication. Considering respective merits and disadvantages at simulated annealing algorithm and genetic algorithm, we combine them .By using simulated annealing algorithm(SA) to mitigate the stress of culling in genetic algorithm(GA) and utilizing the Boltzmann mechanism of simulated annealing algorithm to control the acceptance of the individuals generated by copulated and varied, a CDMA multi-user detection method based on genetic annealing algorithm is adopted. The computer simulative experiment is conducted. Satisfying result is got after analyze the simulation result. The result proved that the use of this algorithm will improve the performance of system and that the performance of system is better than the decorrelating multi-user detection. Experiments proved its efficiency.The following is what I have done in this paper:1.The paper systematically introduced the fundamental knowledge of Spread Spectrum Communication in CDMA , introduced multi-user detection technology in CDMA, and analysed advantages and flaws of multiuser detection technology.2.The paper introduced SA and GA. Through the analysis of faults and merits of two algorithms,we combined Genetic Algorithm with Simulated Annealing Algorithm and formed a genetic annealing algorithm and applied to frequence assignment.3.The paper used a multi-user detection based on genetic annealing algorithm.