Research On Intelligence Computation Used Multi-user Detection Algorithm

Multi-user detection (MUD) is the most effective method to solve multiple accessinterference (MAI) problems, and is one of the key techniques in code divisionmultiple access (CDMA) communication system. Based on the conventional detection(CD), it utilizes the MAI of all the users to detect the target user’s information, whichcan decrease MAI and has better near-far resistance (NFR). So the resource offrequency spectrum is fully utilized, and the capacity of the system is enlarged.Because of the computational complexity grows exponentially with the numberof users in system, optimal multi-user detection (OMD) proposed by S.Verdu cannotbe put into practice. Therefore, a variety of suboptimal detections with lowcomputational complexity and excellent performance become the hot topic, especiallythose intelligence computation used detections.In this thesis, intelligence computation and intelligence computation useddetections are lucubrated.The main contribution of this thesis can be summarized as follows:(1) In the beginning, we proposed an improved adaptive clonal selectionalgorithm (IACSA). Then a detection using IACSA is designed. Experiments havebeen carried out to analyze the performance of this detection.(2) To overcome the defect of probability clonal selection particle swarmoptimization (PCSPSO) algorithm in solving discrete optimization problems, animproved algorithm (IPCSPSO) is researched and used to solve MUD problems.Simulation results show that the researched multi-user detection has significantperformance improvement.(3) A greedy strategy based clonal selection shuffled frog leaping algorithm(GCSSFLA) is proposed by applying discrete shuffled frog leaping algorithm (DSFLA)and greedy strategy in clonal selection algorithm (CSA). Then a novel MUD methodbased on GCSSFLA is presented. Simulation results show that the performance of thismethod is similar to that of OMD.