Study On Application Of Intelligence Computation To MC-CDMA Systems Multi-User Detection

Multiuser detection (MUD) problem can be viewed as a combinational optimization problem. Intelligence computation has shown many advantages over conventional optimization algorithm. This thesis is dedicated to the application of Intelligence computation MUD to antenna-diversity assisted MC-CDMA systems. The main works of this paper can be summarized as follows:1. Analyzed principle of MC-CDMA system, established the mathematic model transmission and reception for MC-CDMA system under the gauss channel and the Rayleigh channel.2. Several commonly used suboptimal multiuser detectors are described. Lastly, the simulation results of these suboptimal detectors are given.3. A genetic-Simulated-Annealing algorithm (GA) assisted multiuser detection (MUD) is proposed for the antenna-diversity-aided multicarrier code-division multiple-access (MC-CDMA) systems in frequency-selective fading channel. Two kinds of GA-based individual-selection strategies are investigated with two different cost functions, which are log-likelihood function (LLF) and square phase error, respectively. Simulation analysis shows that: with the same individual-selection strategy, the bit error rate (BER) performance of the phase error cost function is better than log-likelihood function.With the same cost function, the individual-selection strategy based on Pareto optimality criterion has better performance than the linear combing criterion.4.A improved artificial fish swarm algorithm (AFSA) assisted multi-user detection (MUD) is proposed for the receive antenna diversity aided multi-carrier code-division multiple-access (MC-CDMA) systems in frequency-selective fading channel. The individuals associated with the AFSA are selected based on the concept of Pareto optimality, which uses the information from the antennas independently. Simulation results showed that: with the similar cost function, the strategy of AFSA -MUD has much better bit error rate (BER) performance than the GA-MUD.5. A genetic-Simulated-Annealing algorithm (GA) assisted multiuser detection (MUD) is proposed for the single antenna with two different cost functions aided multicarrier code-division multiple-access (MC-CDMA) systems in frequency-selective fading channel. The two different cost functions are log-likelihood function (LLF) and square phase error, respectively. Simulation analysis shows that: the strategy of GA-MUD based on the single antenna with two different cost functions has much better bit error rate (BER) performance than the GA-MUD based on the single antenna with single cost functions. Simulation results showed that: with the low computation complexity, the bit error rate (BER) performance of the strategy of GA-MUD based on the single antenna with two different cost functions close to the strategy of GA-MUD based on the double antenna with the single cost functions.