一种基于类内聚集度和类间离散度的特征集提取方法
|
摘要
雷达辐射源信号识别是雷达对抗信号处理中的关键技术,特征提取和选择是其中的重要一环。由于受噪声等多种因素影响,信号最重要的特征常常难以发现,特征提取一般采用经验式的方法。为消除特征提取的主观性和提高准确识别率,给出了类内聚集度和类间离散度的定义,并提出了一种基于类内聚集度和类间离散度的特征集评价准则。基于该准则,采用特征排序选择的搜索策略,可以从雷达辐射源信号原始特征集中挑选出品质优良的特征子集,从而实现对雷达辐射源信号的准确识别。
The recognition of radar emitter signals is a key technology in signal processing of radar countermeasures,and feature extraction and selection is a important part of that.It's difficult to find the most notable features of signals due to noise and so many other factors.So feature extraction generally uses empirical method.In order to eliminate the subjectivity of feature extraction and improve the accuracy of recognition,Within-class aggregation and between-class scatter are defined,and a feature set evaluation criteria based on within-class aggregation and between-class scatter is proposed.Based on this criterion,the good-quality feature subsets can be selected from original signal feature sets using the search strategy of sorted feature selection,and the radar emitter signal can be recognized accurately.
The recognition of radar emitter signals is a key technology in signal processing of radar countermeasures,and feature extraction and selection is a important part of that.It's difficult to find the most notable features of signals due to noise and so many other factors.So feature extraction generally uses empirical method.In order to eliminate the subjectivity of feature extraction and improve the accuracy of recognition,Within-class aggregation and between-class scatter are defined,and a feature set evaluation criteria based on within-class aggregation and between-class scatter is proposed.Based on this criterion,the good-quality feature subsets can be selected from original signal feature sets using the search strategy of sorted feature selection,and the radar emitter signal can be recognized accurately.
引文
[1]张葛祥,荣海娜,金炜东.基于小波包变换和特征选择的雷达辐射源信号识别[J].电路与系统学报,2006,11(6):46-49.
[2]张静远,张冰,蒋兴舟.基于小波变换的特征提取方法分析[J].信号处理,2000,16(2):156-162.
[3]钱涛,马健,刘凯,等.基于连通检测和区域扩展的多重聚类方法研究[J].航天电子对抗,2016,32(5):36-39.
[4]杨建,周涛,何梓昂.一种有效的雷达信号快速识别方法[J].电讯技术,2017,57(4):418-424.
[5]普运伟,金炜东,胡来招.基于瞬时频率二次特征提取的辐射源信号分类[J].西南交通大学学报,2007,42(3):373-379.
[6]沈家煌,黄建冲,朱永成.特征参数匹配法在雷达信号识别中的应用[J].航天电子对抗,2017,33(2):9-14.
[7]吴振强,常硕,张国毅.基于信号特征综合处理的雷达辐射源识别[J].科学技术与工程,2015,15(25):12-17.
[8]郭亚琴,王正群.基于类内类间离散度的分类器设计方法[J].信息技术,2010,5(1):35-37.
[2]张静远,张冰,蒋兴舟.基于小波变换的特征提取方法分析[J].信号处理,2000,16(2):156-162.
[3]钱涛,马健,刘凯,等.基于连通检测和区域扩展的多重聚类方法研究[J].航天电子对抗,2016,32(5):36-39.
[4]杨建,周涛,何梓昂.一种有效的雷达信号快速识别方法[J].电讯技术,2017,57(4):418-424.
[5]普运伟,金炜东,胡来招.基于瞬时频率二次特征提取的辐射源信号分类[J].西南交通大学学报,2007,42(3):373-379.
[6]沈家煌,黄建冲,朱永成.特征参数匹配法在雷达信号识别中的应用[J].航天电子对抗,2017,33(2):9-14.
[7]吴振强,常硕,张国毅.基于信号特征综合处理的雷达辐射源识别[J].科学技术与工程,2015,15(25):12-17.
[8]郭亚琴,王正群.基于类内类间离散度的分类器设计方法[J].信息技术,2010,5(1):35-37.