复杂体制雷达辐射源信号特征评价
|
摘要
随着电子对抗的日趋激烈,雷达对抗信号处理产生了许多亟待解决的理论问题。基于常规五参数特征(载频RF、脉冲到达时间TOA、脉冲幅度PA、脉冲宽度PW、脉冲到达方向DOA)进行分选识别已经不能适应含有各种复杂体制雷达(如脉内调频、相位编码、频率捷变、PRI捷变等)的现代电子战信号环境。电子方面的专家对雷达辐射源信号的时域、频域等方面进行了深入的研究,提出了许多雷达辐射源信号特征。然而,由于雷达辐射源信号特征评价体系方面的工作尚未完善,因此,导致已提出的一些特征不能更好的应用到实际的环境中。基于此,本文通过构造雷达辐射源信号的评价指标,运用综合评价模型对雷达辐射源信号特征进行综合评价。
1、分析了目前雷达辐射源信号特征评价研究成果,根据其中存在的一些问题,建立了雷达辐射源信号特征评价模型。
2、阐述和分析了小波包特征、瞬时频率级联特征、时频原子特征的提取算法,其中针对小波包特征维数高,个别特征受噪声干扰严重以及特征有冗余的问题,基于ReliefF算法选择分选识别能力较强的3个特征,从而实现特征的降维和信号正确识别率的提高。通过对线性调频脉冲信号(LFM)、常规脉冲信号(CW)、非线性调频脉冲信号(NLFM)、二相编码信号(BPSK)、四相编码信号(QPSK)和二频率编码信号(BFSK)的调制参数的改变来验证特征对信号调制参数的稳定性并将其作为评价指标之一。
3、从特征的复杂度、基于互信息的可分性和信号的识别率三个角度构建雷达辐射源信号的特征评价指标体系,给出每个特征评价指标的理论模型和实现算法。其中包括时间复杂度指标、空间复杂度指标(包括类内类间距离、交叠域体积、特征效率、最小生成树边界复杂度、类外和类内最近邻平均距离比、基于LP的非线性和基于1NN的非线性)、基于互信息的特征可分性指标和信号识别率的置信区间以及识别率的稳定性指标,通过仿真实验得到每个特征的各项指标的数据。
4、建立了基于模糊层次多目标决策的雷达辐射源信号特征综合评价模型。将文章中提到的各项指标进行综合从而较为全面和准确地评价信号特征的性能,为添加新的特征提供一定的理论依据。
本文在国家自然科学基金《复杂体制雷达辐射源信号特征评价机制》(No.60971103)的资助下开展的。
As the electronic warfare becoming more and more fierce, a number of theoretical problems brought by the signal processing of radar countermeasures need to be solved. Sub-selection identification based on the conventional five parameters (carrier frequency of the RF, pulse time of arrival TO A, the PA of the pulse amplitude, pulse width PW, pulse direction of arrival DOA) can not adapt to modern electronic warfare signal environments with a variety of complex system radars (such as pulse frequency modulation, phase encoding, frequency agility, PRI agility, etc.). Electronic experts have made a lot of research on radar emitter signals in time domain and frequency domain, and many signal features of radar emitter have been put forward. However, the evaluation system of the radar emitter signals'features was not perfect, therefore, some of the features can't been well applied to practical environment. Based on this, this article constructs the evaluation model of radar emitter signals feature, and uses the evaluation model to make a comprehensive evaluation for the radar emitter signal features.
1. Evaluate the status quo of radar emitter signal feature at home and abroad, and point out the problems need to be solved, and at last construct a emitter signal feature evaluation model.
2. Analyse the extraction algorithm based on wavelet packet features, the instantaneous frequency derived features and time-frequency atoms potential modulation features respectively, in view of the high dimension of the wavelet packet features, some of the dimensional feature being interrupted serious by noise and the redundance of the features, the ReliefF algorithm was used to select the sorting ability to identify strong features in order to achieve the reduction of the features of dimensionality and the improvement of the correct recognition rate. The radar emitter signal(Conventional pulse signal, the chirp signal, non-linear FM pulse, two-phase encoded signals, four-phase encoded signals, the modulation parameters of the two frequency encoded signals) modulation parameters was changed in order to verify the stability of the features of the signal parameters. It was used as one of the evaluation index.
3. The features evaluation index system of the emitter signal was established from the features of complexity, separability, and the recognition rate of the signal, given the theoretical model of each feature evaluation algorithm. It includes the time complexity index, index of spatial complexity (including class within the class of the distance between the overlapping domain size feature efficiency, minimum spanning tree variable boundary, outside of class and class nearest neighbor distance, based on the LP, nonlinear1NN-based non-linear), the stability index based on the mutual information of the reparability index, signal recognition rate of the confidence interval and the recognition rate, and the data of all indicators was simulated in the experiment.
4. A radar emitter signal features of the multi-objective decision-making based on fuzzy hierarchy evaluation model is established. The indicators mentioned in the article are integrated in order to assess the performance of the signal features more comprehensively and accurately. This model can provide a theoretical basis for the adding of the other new features.
This work is funded by "the National Natural Science Foundation of the complex system of radar emitter signal features evaluation mechanism "(No.60971103).
1、分析了目前雷达辐射源信号特征评价研究成果,根据其中存在的一些问题,建立了雷达辐射源信号特征评价模型。
2、阐述和分析了小波包特征、瞬时频率级联特征、时频原子特征的提取算法,其中针对小波包特征维数高,个别特征受噪声干扰严重以及特征有冗余的问题,基于ReliefF算法选择分选识别能力较强的3个特征,从而实现特征的降维和信号正确识别率的提高。通过对线性调频脉冲信号(LFM)、常规脉冲信号(CW)、非线性调频脉冲信号(NLFM)、二相编码信号(BPSK)、四相编码信号(QPSK)和二频率编码信号(BFSK)的调制参数的改变来验证特征对信号调制参数的稳定性并将其作为评价指标之一。
3、从特征的复杂度、基于互信息的可分性和信号的识别率三个角度构建雷达辐射源信号的特征评价指标体系,给出每个特征评价指标的理论模型和实现算法。其中包括时间复杂度指标、空间复杂度指标(包括类内类间距离、交叠域体积、特征效率、最小生成树边界复杂度、类外和类内最近邻平均距离比、基于LP的非线性和基于1NN的非线性)、基于互信息的特征可分性指标和信号识别率的置信区间以及识别率的稳定性指标,通过仿真实验得到每个特征的各项指标的数据。
4、建立了基于模糊层次多目标决策的雷达辐射源信号特征综合评价模型。将文章中提到的各项指标进行综合从而较为全面和准确地评价信号特征的性能,为添加新的特征提供一定的理论依据。
本文在国家自然科学基金《复杂体制雷达辐射源信号特征评价机制》(No.60971103)的资助下开展的。
As the electronic warfare becoming more and more fierce, a number of theoretical problems brought by the signal processing of radar countermeasures need to be solved. Sub-selection identification based on the conventional five parameters (carrier frequency of the RF, pulse time of arrival TO A, the PA of the pulse amplitude, pulse width PW, pulse direction of arrival DOA) can not adapt to modern electronic warfare signal environments with a variety of complex system radars (such as pulse frequency modulation, phase encoding, frequency agility, PRI agility, etc.). Electronic experts have made a lot of research on radar emitter signals in time domain and frequency domain, and many signal features of radar emitter have been put forward. However, the evaluation system of the radar emitter signals'features was not perfect, therefore, some of the features can't been well applied to practical environment. Based on this, this article constructs the evaluation model of radar emitter signals feature, and uses the evaluation model to make a comprehensive evaluation for the radar emitter signal features.
1. Evaluate the status quo of radar emitter signal feature at home and abroad, and point out the problems need to be solved, and at last construct a emitter signal feature evaluation model.
2. Analyse the extraction algorithm based on wavelet packet features, the instantaneous frequency derived features and time-frequency atoms potential modulation features respectively, in view of the high dimension of the wavelet packet features, some of the dimensional feature being interrupted serious by noise and the redundance of the features, the ReliefF algorithm was used to select the sorting ability to identify strong features in order to achieve the reduction of the features of dimensionality and the improvement of the correct recognition rate. The radar emitter signal(Conventional pulse signal, the chirp signal, non-linear FM pulse, two-phase encoded signals, four-phase encoded signals, the modulation parameters of the two frequency encoded signals) modulation parameters was changed in order to verify the stability of the features of the signal parameters. It was used as one of the evaluation index.
3. The features evaluation index system of the emitter signal was established from the features of complexity, separability, and the recognition rate of the signal, given the theoretical model of each feature evaluation algorithm. It includes the time complexity index, index of spatial complexity (including class within the class of the distance between the overlapping domain size feature efficiency, minimum spanning tree variable boundary, outside of class and class nearest neighbor distance, based on the LP, nonlinear1NN-based non-linear), the stability index based on the mutual information of the reparability index, signal recognition rate of the confidence interval and the recognition rate, and the data of all indicators was simulated in the experiment.
4. A radar emitter signal features of the multi-objective decision-making based on fuzzy hierarchy evaluation model is established. The indicators mentioned in the article are integrated in order to assess the performance of the signal features more comprehensively and accurately. This model can provide a theoretical basis for the adding of the other new features.
This work is funded by "the National Natural Science Foundation of the complex system of radar emitter signal features evaluation mechanism "(No.60971103).
引文
[1]候印鸣.综合电子战—现代战争的杀手锏.北京：国防工业出版社,2000.
[2]Schroer R. Electronic warfare. IEEE Aerospace Electronic Systems Magazine.2003,18(7): 49-54.
[3]胡来招.雷达侦察接收机.北京：国防工业出版社,2000.
[4]张国柱.雷达辐射源识别技术研究.长沙：国防科学技术大学博士学位论文.2005.
[5]陈韬伟,金炜东.雷达辐射源信号符号化脉内特征提取方法.数据采集与处理,2008,23(5)：521-526.
[6]Zhu Ming, Pu Yunwei, Jin Weidong, Hu Laizhao. A Time-Frequency Atom Approach to Radar Emitter Signal Feature Extraction. Proceedings of IEEE International Conf. on Communications, Circuits and Systems, 2006, pp.615-619.
[7]Zhang Ge Xiang, Marian Gheorghe, Wu Chao Zhong, A quantum-inspired evolutionary algorithm based on P systems for knapsack problem, Fundamenta Informaticae, IOS Press, 2008,87(1):93-116.
[8]许学,陈建.基于小子样Bootstrap法的雷达辐射源特征分选稳定性分析.成都电子机械高等专科学校学报,2011年第2期.
[9]张贤达,保铮.非平稳信号分析与处理.北京：国防工业出版社,1998.
[10]向敬成,张明友.雷达系统.北京：电子工业出版社,2001.
[11]毕大平,董晖,姜秋喜.基于瞬时频率的脉内调制识别技术.电子对抗技术.2005.
[12]Rong Haina, Jin Weidong, Zhang Cuifang. Application of Support Vector Machines to Pulse Repetition Interval Modulation Recognition. Proceedings of International Conference on ITS Telecommunication,2006; 1187-1190.
[13]Yu Zhibin, Jin Weidong, Zhang Gexiang. Multi-component LFM Radar Emitter Signal Detection Based on LWD. Proceedings of the 7th World Congress on Intelligent Control and Automation,2008, vol.6, pp.4463-4467.
[14]Chen Taowei, Jin Weidong. Feature extraction of radar emitter signals based on symbolic time series analysis. Proceedings of International Conference on Wavelet Analysis and Pattern Recognition, 2007, pp.1277-1282.
[15]张葛祥.雷达辐射源信号智能识别方法研究.成都：西南交通大学博士学位论文.2005.
[16]普运伟,金炜东,胡来招.基于瞬时二次特征提取的辐射源信号分类.西南交通大学学报.2007,42(3)：373-379.
[17]程吉祥,张葛祥,唐承志.复杂体制雷达辐射源信号时频原子特征提取方法.西安交通大学学报.2010,44(4)：108-113.
[18]丁幼亮,李爱群,缪长青.基于小波包能量谱的结构损伤预警方法研究[J].工程力学,2006,23(8)：29-38.
[19]Hu X G, Wang F, Zhao H L, et al. The mechanical fault diagnosis for HV breakers on the wavelet packet analysis [A]. Proceedings of Instrumentation and Measurement Technology Conference [C].2003.415-419.
[20]Ye Z G, Wu B, Sadeghian A. Current signature analysis of induction motor mechanical fault by wavelet packet decomposition [J]. IEEE Transaction on Industrial Electronics, 2003,50(6):1217-1228.
[21]刘庆云,陆飞飞,朱伟强等.辐射源细微特征用于个体识别的可行性分析.航天电子对抗.2008,24(2)：40-42.
[22]余志斌,金炜东.基于小波脊频特征的雷达辐射源信号识别.西南交通大学学报,2009.
[23]王晓华.雷达信号脉内时频分析的一种新方法.舰船电子对抗.2006,29(6)：67-69.
[24]Lunden J, Koivunen V. Automatic radar waveform recognition. IEEE Journal of Selected Topics in Signal Processing. 2007,1 (1):124-136.
[25]朱明.复杂体制雷达辐射源信号时频原子特征研究.成都：西南交通大学博士学位论文.2008.
[26]Cui J, Wong W, Mann S. Time-frequency analysis of visual evoked potentials by means of matching pursuit with chirplet atoms. Proceedings of the International Conference on Engineering in Medicine and Biology Society, 2004. IEEE:267-270.
[27]Lopez R G, Grajal J, Yeste O O. Atomic decomposition-based radar complex signal interception. Proceedings of the IEE Conference on Radar, Sonar and Navigation, 2003. IEEE:323-331.
[28]SUN Y J. Iterative RELIEF for feature weighting:algorithms, theories, and applications [J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2007,29 (6): 1035-1051.
[29]ZHANG JJ, LIN H, ZHAO M G. A Fast algorithm for hand gesture recognition using relief[C]. China:Sixth International Conference on Fuzzy Systems and Knowledge Discovery, 2009:8-12.
[30]张丽新.高维数据的特征选择及基于特征选择的集成学习研究[D].北京：清华大学,2004.
[31]KIRA K, RENDLL L A. The feature selection problem:traditional methods and a new algorithm[C]. California:Ninth National conference on Aritifical Intelligence, 1992: 129-134.
[32]MARKO R S, IGOR K. Theoretical and empirical analysis of relieff and rreliefF[J]. Machine Learning, 2003,53:23-69.
[33]J.H.Friedman and L.C.Rafsky,"Multivariate Generalizations of the Wald-Wolfowitz and Smirnov Two-Sample Tests,"The Annals of Statistics,vol.7,no.4,pp.697-717,1979.
[34]T.K.Ho and M.Basu,"Measuring the Complexity of Classification Problems,"Proc,15th Int`l Conf.Pattern Recognition,vol.2,pp.4347,2000.
[35]Burrus C.S.小波与小波变换导论.北京：机械工业出版社,2005.pp.170-173.
[36]M.Li and P.Vitanyi,An Introduction to Kolmogorov Complexity and Its Applications. Springer-Verlag,1993.
[37]J.M.Maciejowski,"ModelDiscrimination Using an Algorithmic Information Criterion ", Automatica,vol.15,pp.579-593,1979.
[38]N.Wyse,R.Dubes,and A.K.Jain,"A Critical Evaluation of Intrinsic Dimensionality Algorithms",PatternRecognitioninPractice,E.S.GelsemaandL.N.Kanal,eds.,North-Holland, pp.415-425,1-985.
[39]李峥.情报综合中相关门限的统计分析.电子对抗技术.2001,16(6)：31-35.
[40]电子情报研究报告——信息战及其装备技术发展研究.电子工业部科学技术情报研究所,1998.
[41]孙即详.现代模式识别[M].国防科技大学出版社,2002.
[42]何峻,卢再奇,付强.ATR算法稳定性评估方法.现代雷达,2006,28(9)：59-61.
[43]Bar Shalom Y, TseE. Tracking in a cluttered environment with probabilistic data association[J]. Automatic,1975,11 (9):451-460.
[44]Li X R. Engineer&sguide to variable structure multiple model estimation for tracking [C] //Multi target, Multi sensor T racking:Applications and Advances, A rtechHouse, Boston,2000.
[45]孙红才.网络层次分析法与决策科学[M].北京：国防工业出版社,2011.
[46]SAATYTL. The Analytic Hierarchy Process. New York:Mc Graw2 Hill,1980.
[47]陈守煜.工程模糊集理论与应用.北京：国防工业出版社,1998.
[48]何平,数理统计与多元统计.成都：西南交通大学出版社,2004.
[49]B.Efron.R.J.Tibshirani.Bootstrap methods for standard errors,confidence interval,and other measures of statistical accuracy.Statistical Science,1986:54-77.
[50]刘伟,自动目标识别系统效能评估方法研究.国防科学科技大学,2007.
[51]贾正源,赵亮.基于熵权未确知测得模型的电能质量综合评价[J].电力系统保护与控 制,2010,38(15)：33-37.
[52]雷新.多因素模糊评判在雷达辐射源识别中的应用[D].北京邮电大学.
[2]Schroer R. Electronic warfare. IEEE Aerospace Electronic Systems Magazine.2003,18(7): 49-54.
[3]胡来招.雷达侦察接收机.北京：国防工业出版社,2000.
[4]张国柱.雷达辐射源识别技术研究.长沙：国防科学技术大学博士学位论文.2005.
[5]陈韬伟,金炜东.雷达辐射源信号符号化脉内特征提取方法.数据采集与处理,2008,23(5)：521-526.
[6]Zhu Ming, Pu Yunwei, Jin Weidong, Hu Laizhao. A Time-Frequency Atom Approach to Radar Emitter Signal Feature Extraction. Proceedings of IEEE International Conf. on Communications, Circuits and Systems, 2006, pp.615-619.
[7]Zhang Ge Xiang, Marian Gheorghe, Wu Chao Zhong, A quantum-inspired evolutionary algorithm based on P systems for knapsack problem, Fundamenta Informaticae, IOS Press, 2008,87(1):93-116.
[8]许学,陈建.基于小子样Bootstrap法的雷达辐射源特征分选稳定性分析.成都电子机械高等专科学校学报,2011年第2期.
[9]张贤达,保铮.非平稳信号分析与处理.北京：国防工业出版社,1998.
[10]向敬成,张明友.雷达系统.北京：电子工业出版社,2001.
[11]毕大平,董晖,姜秋喜.基于瞬时频率的脉内调制识别技术.电子对抗技术.2005.
[12]Rong Haina, Jin Weidong, Zhang Cuifang. Application of Support Vector Machines to Pulse Repetition Interval Modulation Recognition. Proceedings of International Conference on ITS Telecommunication,2006; 1187-1190.
[13]Yu Zhibin, Jin Weidong, Zhang Gexiang. Multi-component LFM Radar Emitter Signal Detection Based on LWD. Proceedings of the 7th World Congress on Intelligent Control and Automation,2008, vol.6, pp.4463-4467.
[14]Chen Taowei, Jin Weidong. Feature extraction of radar emitter signals based on symbolic time series analysis. Proceedings of International Conference on Wavelet Analysis and Pattern Recognition, 2007, pp.1277-1282.
[15]张葛祥.雷达辐射源信号智能识别方法研究.成都：西南交通大学博士学位论文.2005.
[16]普运伟,金炜东,胡来招.基于瞬时二次特征提取的辐射源信号分类.西南交通大学学报.2007,42(3)：373-379.
[17]程吉祥,张葛祥,唐承志.复杂体制雷达辐射源信号时频原子特征提取方法.西安交通大学学报.2010,44(4)：108-113.
[18]丁幼亮,李爱群,缪长青.基于小波包能量谱的结构损伤预警方法研究[J].工程力学,2006,23(8)：29-38.
[19]Hu X G, Wang F, Zhao H L, et al. The mechanical fault diagnosis for HV breakers on the wavelet packet analysis [A]. Proceedings of Instrumentation and Measurement Technology Conference [C].2003.415-419.
[20]Ye Z G, Wu B, Sadeghian A. Current signature analysis of induction motor mechanical fault by wavelet packet decomposition [J]. IEEE Transaction on Industrial Electronics, 2003,50(6):1217-1228.
[21]刘庆云,陆飞飞,朱伟强等.辐射源细微特征用于个体识别的可行性分析.航天电子对抗.2008,24(2)：40-42.
[22]余志斌,金炜东.基于小波脊频特征的雷达辐射源信号识别.西南交通大学学报,2009.
[23]王晓华.雷达信号脉内时频分析的一种新方法.舰船电子对抗.2006,29(6)：67-69.
[24]Lunden J, Koivunen V. Automatic radar waveform recognition. IEEE Journal of Selected Topics in Signal Processing. 2007,1 (1):124-136.
[25]朱明.复杂体制雷达辐射源信号时频原子特征研究.成都：西南交通大学博士学位论文.2008.
[26]Cui J, Wong W, Mann S. Time-frequency analysis of visual evoked potentials by means of matching pursuit with chirplet atoms. Proceedings of the International Conference on Engineering in Medicine and Biology Society, 2004. IEEE:267-270.
[27]Lopez R G, Grajal J, Yeste O O. Atomic decomposition-based radar complex signal interception. Proceedings of the IEE Conference on Radar, Sonar and Navigation, 2003. IEEE:323-331.
[28]SUN Y J. Iterative RELIEF for feature weighting:algorithms, theories, and applications [J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2007,29 (6): 1035-1051.
[29]ZHANG JJ, LIN H, ZHAO M G. A Fast algorithm for hand gesture recognition using relief[C]. China:Sixth International Conference on Fuzzy Systems and Knowledge Discovery, 2009:8-12.
[30]张丽新.高维数据的特征选择及基于特征选择的集成学习研究[D].北京：清华大学,2004.
[31]KIRA K, RENDLL L A. The feature selection problem:traditional methods and a new algorithm[C]. California:Ninth National conference on Aritifical Intelligence, 1992: 129-134.
[32]MARKO R S, IGOR K. Theoretical and empirical analysis of relieff and rreliefF[J]. Machine Learning, 2003,53:23-69.
[33]J.H.Friedman and L.C.Rafsky,"Multivariate Generalizations of the Wald-Wolfowitz and Smirnov Two-Sample Tests,"The Annals of Statistics,vol.7,no.4,pp.697-717,1979.
[34]T.K.Ho and M.Basu,"Measuring the Complexity of Classification Problems,"Proc,15th Int`l Conf.Pattern Recognition,vol.2,pp.4347,2000.
[35]Burrus C.S.小波与小波变换导论.北京：机械工业出版社,2005.pp.170-173.
[36]M.Li and P.Vitanyi,An Introduction to Kolmogorov Complexity and Its Applications. Springer-Verlag,1993.
[37]J.M.Maciejowski,"ModelDiscrimination Using an Algorithmic Information Criterion ", Automatica,vol.15,pp.579-593,1979.
[38]N.Wyse,R.Dubes,and A.K.Jain,"A Critical Evaluation of Intrinsic Dimensionality Algorithms",PatternRecognitioninPractice,E.S.GelsemaandL.N.Kanal,eds.,North-Holland, pp.415-425,1-985.
[39]李峥.情报综合中相关门限的统计分析.电子对抗技术.2001,16(6)：31-35.
[40]电子情报研究报告——信息战及其装备技术发展研究.电子工业部科学技术情报研究所,1998.
[41]孙即详.现代模式识别[M].国防科技大学出版社,2002.
[42]何峻,卢再奇,付强.ATR算法稳定性评估方法.现代雷达,2006,28(9)：59-61.
[43]Bar Shalom Y, TseE. Tracking in a cluttered environment with probabilistic data association[J]. Automatic,1975,11 (9):451-460.
[44]Li X R. Engineer&sguide to variable structure multiple model estimation for tracking [C] //Multi target, Multi sensor T racking:Applications and Advances, A rtechHouse, Boston,2000.
[45]孙红才.网络层次分析法与决策科学[M].北京：国防工业出版社,2011.
[46]SAATYTL. The Analytic Hierarchy Process. New York:Mc Graw2 Hill,1980.
[47]陈守煜.工程模糊集理论与应用.北京：国防工业出版社,1998.
[48]何平,数理统计与多元统计.成都：西南交通大学出版社,2004.
[49]B.Efron.R.J.Tibshirani.Bootstrap methods for standard errors,confidence interval,and other measures of statistical accuracy.Statistical Science,1986:54-77.
[50]刘伟,自动目标识别系统效能评估方法研究.国防科学科技大学,2007.
[51]贾正源,赵亮.基于熵权未确知测得模型的电能质量综合评价[J].电力系统保护与控 制,2010,38(15)：33-37.
[52]雷新.多因素模糊评判在雷达辐射源识别中的应用[D].北京邮电大学.