| Code Division Multiple Access (CDMA) is the main technology of the third generation mobile telecommunication, it uses different address codes to distinguish different users. But in practical the address codes can not completely orthogonal, so it causes multiple access interference (MAI), which can impress the performance of the CDMA system. Multiuser detection (MUD) is the key technology to remove the MAI. The optimal multiuser detection can get perfect performance but its computational complexity grows exponentially with the numbers of users, so it can not be put into practice. At the same time we find the optimal detection is a typical NP-complete problem, and the optimal algorithms are very efficient in solving such problems. The main work is summarized as follows:1) The principle of MUD, MUD system and some classical suboptimal MUD based on the research on spread spectrum and synchronize CDMA system are being researching in this article, and also the performs of some classical suboptimal MUD are analyzed and compared.2) Putting Simple Genetic Algorithm (SGA) into the MUD system and using the good global search ability of GA to find the optimal value, and also compare the performance with the GA-MUD and the classical MUD above.3) Two types of Hybrid Genetic Algorithms (HGA) MUD are main researched in this article, they are MUD based on Simulated Annealing Genetic Algorithm (SAGA) and Hopfield Neural Network Genetic Algorithm (HNNGA), in the HNNGA we use serial and parallel working patterns. On one hand HGA uses GA to provide a global optimal value, on the other hand HGA uses SA or HNN to find the local optimal value and to improve the search rate. The computer simulation shows that these two types of HGA-MUD can get better performs than SGA-MUD in removing MAI and resisting the near-far effects, at the same time, the calculate times of HGA-MUD is much less than SGA-MUD, so SAGA-MUD and HNNGA-MUD are reliable and effective modern MUD techniques. |
PDF Full Text Request