摘要
研究一种基于遗传算法(GA)的模糊聚类方法,即将遗传算法得到的聚类中心作为模糊C-均值(FCM)聚类算法初值,这样既可以克服FCM算法对初始中心敏感的缺点,也可以解决遗传算法只能找到近似解的问题。将算法用于通信信号的星座聚类,根据聚类有效性函数自适应地确定聚类中心,并完成信号类型的识别。仿真实验证明,当存在较小的定时误差时,算法对PSK和QAM信号仍然是有效的。
A fuzzy clustering arithmetic based on Genetic Algorithm （GA） was proposed. Genetic algorithms instructed to choose the initial cluster centers firstly, and clustering was processed by Fuzzy C-Mean Clustering （FCM）. It not only overcomes the sensitivity to initial centers as FCM, but also solves the problem of approximate solution as Genetic Algorithm. This method was applied to constellation clustering and modulation recognition. The simulation results show that the method is feasible to PSK and QAM signal even when timing error exists.
出处
《计算机应用》
CSCD
北大核心
2008年第5期1197-1199,共3页
journal of Computer Applications
基金
总装备部重点科研基金项目(6130320)
关键词
遗传算法
模糊C-均值
星座聚类
Genetic Algorithm（GA）
Fuzzy C-Mean（FCM）
constellation clustering
