基于聚类分析的常规通信信号自动分选方法

Automatic sorting method of conventional communication signals based on clustering analysis

为了从干扰环境中自动分选出常规通信信号,提出了一种基于聚类分析的信号分选方法,它主要包括测量集分割和信号跟踪两部分。测量集分割用于对频域检测和测向得到的测量集进行识别,从中获得对该测量集有贡献的窄带信号的特征参数;信号跟踪就是根据常规通信信号的特点,对截获信号进行序贯聚类,剔除干扰信号。实验结果表明,测量集分割方法的识别正确率在94%以上,能够准确估计各窄带信号的特征参数;所提出的分选方法能够从干扰环境中正确分选出常规通信信号。

To automatically sort conventional communication signals from interference surroundings,this paper proposed a clustering analysis based sorting method.The sorting method consisted of the following two key parts,measurement set partition and signal tracking.The first part identified narrowband signals included in measurement set which originated from the result of frequency detection and direction finding,and estimated the basic characteristic parameters of each signal.According to the characteristics of conventional communication signals,the second one was able to perform sequential clustering on the cha-racteristic parameters of all intercepted signals and delete interference signals.The simulation results show that the correct identification rate of the proposed partition method is better than 94% and this method can accurately estimate the characteristic parameters of each intercepted narrowband signal.The proposed sorting method is able to correctly sort conventional communication signals from interference surroundings.