Quantum Clone Genetic Algorithm Based Multi-User Detecter

With the progress and the development of the society, the mobile communication technology of DS-CDMA also developed saltantly. So far, it has been already developed to the third generation (3G), it brings more convenient and shoutcut service to the work and the daily life of the people. However, with the continual enriched content of the multi-media operation and high speed data operation, it demands the mobile communication of the new generation must have the bigger system capability.At the present, with the increasing of the user's number, the interference among users gets more and more severiouly in DS-CDMA system. Here, the multi-address interference is the uppermost interference, it is a villagal interference among users. Its existence restricts seriously the performance and capability of the mobile communication system. It compels investiqators search a technology of eliminate multi-address interference (MAI). Sequently, investiqators put forward multi-user detection. The basic idea of multi-user detection is that it takes the MAI as a useful user information, other than deal with it simplily as a interfere noise. It makes use of the relevancy among users to carry through unite detection at the most extent, it can eliminate the MAI validly or decrease the MAI to the minimum. Therefore, it can achieve the aim of enhance the performance and the capability of the system.Since the quantum clone genetic algorithm was added immunity clone strategy, it had the characteristics of avoiding blindness and randomness of the algorithm, could avoid prematurity and accelerate converquence, but the enhance of the algorithm's performance was extending space search range and increasing calculate quantity for cost, moreover, the algorithm adopted the variation stratege of quantum rotate gate according to the quantum composition features and the theory of quantum transition, which made the comfirm of the global search direction of quantum gate existed blindness.Therefore, against these defects, at the foundation of quantum clone genetic algorithm which was given by Li Yangyang, in the early state of the evolution, adopting the niche cooperate with evolution strategy to inilialize quantum conoly, which is convenient to the optimal search of the optimal chromosome; in the metaphase of the evolution, inducting quantum with all cross, which is carried through the information transfer in the whole colony and avoids getting into local optimal value, accelerating the converquence of the algorithm, while, the algorithm uses adaptive quantum rotate gate strategy, which can accleralate the search of the optimal solution; in the later stage of the evolution, in order to avoid the prematurity and evolution stagnancy, adopting quantum catastrophe strategy, which uses the all interference cross of the individual and makes the colony search the optimal solution through each direction, therefore, the text give a Improved Quantum Clone Genetic Algorithm (IQCGA).By the evidence of the global converquence and the demonstration of the performance test, the algorithm of IQCGA has the merits of good conoly diversity, strong ability of global search optimal, high efficiency of the search, fast speed of converquence and the proper efficiency of the algorithm. It can be applied to the combined optimal problem.The multi-user detection is a combined optimal question, the text gives a multi-user detection algorithm of DS-CDMA based on IQCGA. And compared with GA-MUD, QGA- MUD and QCGA-MUD by using Matlab to process the simulation, the result of the simulation demonstrates: the bit error, the converquece speed, the resist performance of MAI and the resist performance of"near-far"effect of IQCGA-MUD are all better than GA-MUD, QGA-MUD and QCGA-MUD.