基于模糊KHM聚类的跳频网台分选方法  

A Frequency-hopping Network Station Sorting Method Based on Fuzzy KHM Clustering Algorithm

作　　者：钟兆根 杨芸丞 张立民 ZHONG Zhaogen;YANG Yuncheng;ZHANG Limin(Navy Aviation University,Yantai 264001,China)

机构地区：[1]海军航空大学

出　　处：《电讯技术》2019年第6期699-704,共6页Telecommunication Engineering

基　　金：国家自然科学基金资助项目(61179016);国家自然科学基金重大研究计划(91538201);“泰山学者”建设工程专项基金(ts201511020)

摘　　要：针对聚类分析实现跳频网台分选时,分选结果对初始聚类中心敏感且某些样本点“既可以属于类A也可以属于类B”的问题,提出了一种基于模糊K调和均值(KHM)聚类的跳频网台分选方法。首先利用搜索统计直方图位置法预估聚类数目和聚类中心,减少了算法的迭代次数;然后根据跳频信号的各项参数,应用模糊KHM聚类算法对跳频网台进行分选,有效解决了样本点的隶属度问题;最后通过类内类间距法估计得到准确的聚类数目K、聚类中心位置,大幅提升了聚类算法准确度。仿真结果表明,该算法聚类中心接近实际类中心,分选正确率高,迭代次数少。In order to solve the problem that the sorting result is sensitive to the initial clustering center and some sample points belong to either class A or class B when clustering analysis is used to realize the sorting of frequency-hopping(FH) network stations,a sorting method based on fuzzy K harmonic means(KHM) clustering is proposed.Firstly,the search statistics histogram position method is used to estimate the number of clusters and the cluster center,thus reducing the number of iterations of the algorithm.Then,according to the parameters of the FH signal,the fuzzy KHM clustering algorithm is applied to sort the frequency hopping network,so the problem of the membership degree of the sample points is effectively solved.Finally,the accurate cluster number K and the cluster center position are estimated by the intra-class class spacing method,which makes the accuracy of the clustering algorithm greatly improved.The simulation results show that the clustering center of the algorithm is close to the actual class center,the sorting accuracy is high,and the number of iterations is small.

关 键 词：跳频通信 网台分选 聚类 模糊KHM 

分 类 号：TN974[电子电信—信号与信息处理]