基于多群竞争的并行进化规划多用户检测器  

Multiuser Detector Based On Multi-Subgroup Competition Parallel Evolutionary Programming

作　　者：许良凤[1] 胡敏[1] 林辉[1] 

机构地区：[1]合肥工业大学计算机与信息学院通信系,合肥230009

出　　处：《电子测量与仪器学报》2008年第5期12-15,共4页Journal of Electronic Measurement and Instrumentation

基　　金：安徽省自然科学基金(070412046)项目资助

摘　　要：在CDMA系统的多用户检测器中,求解最佳矢量问题可以转化为在进化规划中求解具有最高适应度函数的问题,利用进化规划能全局寻优的优势,提出了一种基于多群竞争的并行进化规划多用户检测器。在该检测器中,进化在多个不同的子群中并行进行,不同的子群根据其竞争能力使用不同的变异策略,实现种群既具有大范围的搜索能力,也具有局部寻优能力,避免了进化规划算法中的易早熟收敛的问题。通过性能分析和计算机仿真结果表明,这种新的多用户检测器在抗多址干扰的能力和抗远近效应的能力都优于并行进化规划的多用户检测器,并且该检测器的收敛速度明显比并行进化规划检测器快。It is shown that the problem of minimizing the objective function of the optimal multiuser detector can be translated into the problem of finding the highest fitness function in evolutionary programming in multiuser detector of CDMA system. Using the global optimum performance of evolutionary programming, a kind of multiuser detector based on multi-subgroup competition parallel evolutionary programming is proposed. In the detector, evolutions of subgroups are performed in parallel and with different mutation strategies determined by the competition capabilities among subgroups, and then the population can not only search large space but also seek the local optimum, so the premature convergence of evolutionary programming is avoided. Performance analysis and computer simulation results show that the new multiuser detector has better performance than the parallel evolutionary programming multiuser detector in the aspects of resisting multiple-access interference and near-far effect, and converges more quickly than the parallel evolutionary programming muhiuser detector.

关 键 词：CDMA 多用户检测器 进化规划 并行 竞争 

分 类 号：TN914.53[电子电信—通信与信息系统]