改进的近邻函数值准则聚类算法对雷达信号的分选

Sorting Radar Signals by Improved Neighbor Function Value Criterion Clustering Algorithm

针对复杂未知雷达信号的处理,提出了一种改进算法应用于雷达辐射源聚类分选的方法。该方法基于近邻函数值准则聚类算法,并通过引入类调控因子的概念对该聚类算法进行改进,用于对未知雷达信号的聚类分选。仿真实验结果验证了该方法对复杂未知雷达信号分选的有效性和优越性,且能够有效抑制聚类结果“增批”的现象,并能够准确估计出雷达辐射源的数目。

Aimed at the processing of complex and unknown radar signals,an improved algorithm for clustering and sorting of radar emitters.The method is based on the nearest neighbor function value criterion clustering algorithm,and the clustering algorithm is improved by introducing the concept of class regulatory factor,which is used to cluster and sort unknown radar signals.The simulation results verify the effectiveness and superiority of the method for sorting complex unknown radar signals,and can effectively suppress the phenomenon of“increasing batches”of clustering results,and can also accurately estimate the number of radar radiation sources.