摘要
针对BFSN算法需要人工输入参数的缺陷,提出了一种自适应确定参数的SA-BFSN聚类方法。该方法利用同一部雷达数据的分布特性进行聚类。通过确定邻居点和比较相似性以达到分选不同的雷达信号,适用于未知雷达信号的分选。提出了合批的方法和一种分选的联合处理。算法测试表明,该方法对噪声不敏感,能够发现任意形状、大小和密度的聚类。
Since BFSN algorithm requires parameter input by hands,a self-adaptive broad first search neighbors(SA-BFSN) clustering method is proposed to determine the parameters.The method uses the data distribution characteristics of the same radar to carry out the cluster.By determining neighbor points and comparing the their similarity,the sorting of different radar signals is achieved.It is suitable for the sorting of unknown radar signals.A batch and sorting joint processing approach is proposed.Testing results of the algorithm show that the method is not sensitive to noise,and can find the cluster with any shape,size and density.
出处
《现代电子技术》
2012年第13期1-3,6,共4页
Modern Electronics Technique