基于密度聚类与SVM的非法广播信号识别研究

针对非法广播信号的危害,以及传统人工检测效率低的问题,提出一种基于密度聚类与SVM的信号识别模型。首先,采用标准欧式距离对特征信号进行提取;其次,以聚类样本为基础,采用SVM分类器对信号分类;最后,以青海广播电视局中波台整点时刻前后300帧的数据为样本,以静音信号作为评价指标,对信号进行识别。结果表明,在正常信号中加入非法信号后,频谱中有少量的静音信号,且SVM训练时间和识别正确率都要优于传统算法。

Research on illegal broadcast signal recognition based on density clustering and SVM

Aiming at the harm of illegal broadcasting signals and the low efficiency of traditional manual detection,a signal recognition model based on density clustering and SVM is proposed.Firstly,the stan-dard Euclidean distance is used to extract the characteristic signals;secondly,the SVM classifier is used to classify the signals based on clustering samples;finally,the data of 300 frames before and after the whole time in Qinghai Radio and Television Bureau are taken as samples,and the silent signal is taken as evaluation index to identify the signals.The results show that after adding illegal signals to normal signals,there are a small number of silent signals in the spectrum,and the training time and recognition accuracy of SVM are better than traditional algorithms.