基于并行遗传算法的去相关多用户检测器

通过将码分多址通信系统中的最佳多用户检测视为组合优化问题,利用遗传算法能全局寻优的优势,提出了一种基于并行遗传算法的去相关多用户检测器。理论分析和仿真表明:该多用户检测器无论是抗多址干扰的能力还是抗远近效应的能力都明显优于传统的检测器和去相关多用户检测器;由于采用了基于"联姻"策略的并行遗传算法,使这种多用户检测器更易于实时应用和硬件实现。

A Decorrelating Multiuser Detector Based on Parallel Genetic Algorithm

By viewing the optimum multiuser detector in code division multiple access communications systems as a combinational optimum problem, we apply the global optimum performance of genetic algorithms to design a kind of decorrelating multiuser detector based on parallel genetic algorithms(PGADEC-MUD). Theoretical analyses and numerical simulation results have shown that in aspect of multiple-access interference and near-far resistance, the PGADEC-MUD is better than the conventional detector and decorrelating multiuser detector and can be easily implemented by VLSI technology, which benefites from the parallel genetic algorithm.