蒋兵,茅玉龙,曹俊纺

• 1:南京信息工程大学电子与信息工程学院
• 2:中国船舶重工集团公司第七二四研究所

摘要(Abstract)：

针对雷达信号调制类型多样且识别成功率较低的问题,提出一种基于瞬时频率特征提取的雷达信号快速识别新算法。首先利用短时傅里叶变换(STFT)对信号进行特征提取,将提取到的特征进行排列组合,最后把重新组合的特征作为PNN神经网络的输入向量对其进行调制类型的识别。其中,通过改进PNN神经网络的输入输出部分使得最后的输出结果为各组合特征输入向量所得结果的最优值。通过计算机仿真对上述方法进行验证,并分析了PNN神经网络的结构和性能,与BP神经网络进行了比较。仿真结果表明,在雷达信号调制类型识别中应用PNN神经网络能大幅提高其识别效率,并拥有明显优于BP神经网络的分类性能。

关键词(KeyWords)： 雷达信号;短时傅里叶变换;脉内调制;特征选择;概率神经网络;信号识别

Abstract:

Keywords:

基金项目(Foundation):

作者(Author): 蒋兵,茅玉龙,曹俊纺

DOI: 10.16652/j.issn.1004-373x.2018.23.015

参考文献(References)：

• [1]王培培,徐才宏.瞬时频率雷达信号脉内调制识别技术[J].现代电子技术,2016,39(13):30-34.WANG Peipei,XU Caihong.Instantaneous frequency based intrapulse modulation recognition technology of radar signal[J].Modern electronics technique,2016,39(13):30-34.
• [2]章琴,刘以安.一种新的雷达辐射源信号识别方法[J].微型机与应用,2015,34(12):48-53.ZHANG Qin,LIU Yian.A new method of radarsignal recognition[J].Microcomputer&its applications,2015,34(12):48-53.
• [3]DIMITRIOS K,YANNIS M.Prediction in wireless networks by Markov chains[J].IEEE wireless communications,2009,16(2):56-53.
• [4]MCKENNA E,BEACH N.人力资源管理[M].丁凡,译.北京:中信出版社,1998.MCKENNA E,BEACH N.Human resource management[M].DING Fan,Translation.Beijing:CITIC Press,1998.
• [5]张君.宽带雷达信号的脉内调制类型分析工程化算法研究[D].北京:国防科学技术大学,2011.ZHANG Jun.Research on engineering algorithm of intrapulse modulation type analysis ofwideband radar signal[D].Beijing:National University of Defense Technology,2011.
• [6]毛永毅,张颖.非视距传播环境下的AOA定位跟踪算法[J].计算机应用,2011,31(2):317-319.MAO Yongyi,ZHANG Ying.AOA location tracking algorithm in non-line-of-sight propagation environment[J].Journal of computer applications,2011,31(2):317-319.
• [7]王晓峰,邢敏捷,刘歌.基于改进DFT相位差正弦波频率估计[J].电讯技术,2016,56(10):1129-1133.WANG Xiaofeng,XING Minjie,LIU Ge.Improved DFT phase difference sine wave frequency estimation[J].Telecommunication engineering,2016,56(10):1129-1133.
• [8]黄高明,苏国庆,张琪,等.基于神经网络的雷达辐射源智能识别系统[J].雷达科学与技术,2005,3(2):86-90.HUANG Gaoming,SU Guoqing,ZHANG Qi,et al.Radar emitter intelligent identification system based on neural network[J].Radar science and technology,2005,3(2):86-90.
• [9]唐健仁,朱元清,徐庆,等.用神经网络进行雷达辐射源识别研究[J].空军雷达学院学报,2007,21(1):25-27.TANG Jianren,ZHU Yuanqing,XU Qing,et al.Research on radar radiation source identification using neural network[J].Journal of Air Force Radar Academy,2007,21(1):25-27.
• [10]张德丰.Matlab神经网络应用设计[M].北京:机械工业出版社,2011.ZHANG Defeng.Neural network application design based on Matlab[M].Beijing:China Machine Press,2011.
• [11]巴寅亮,王书提,李春兰,等.基于PNN神经网络的电控发动机故障诊断[J].现代电子技术,2016,39(20):146-153.BA Yinliang,WANG Shuti,LI Chunlan,et al.PNN based fault diagnosis of electrically controlled engine[J].Modern electronics technique,2016,39(20):146-153.
• [12]王肖灵,王波,张惠珍,等.GRNN和PNN模型在信用风险评估中的应用[J].数学理论与应用,2015,35(3):101-106.WANG Xiaoling,WANG Bo,ZHANG Huizhen,et al.Application of GRNN and PNN models in credit risk assessment[J].Mathematical theory and applications,2015,35(3):101-106.
• [13]吴昊.雷达信号脉内特征实时分析处理技术研究[D].北京:中国舰船研究院,2016.WU Hao.Research of real-time analysis and processing technique of radar signal intra pulse feature[D].Beijing:China Ship Research and Development Academy,2016.