计算机科学

计算机科学 ›› 2024, Vol. 51 ›› Issue (5): 267-276.doi: 10.11896/jsjkx.230300216

• 计算机网络 • 上一篇    下一篇

基于混合式特征选择的辐射源个体识别

顾楚梅1,2,3, 曹建军1,2, 王保卫3, 徐雨芯1,2,3   

  1. 1 国防科技大学第六十三研究所 南京 210007
    2 国防科技大学大数据与决策实验室 长沙 410073
    3 南京信息工程大学计算机学院网络空间安全学院 南京 210044
  • 收稿日期:2023-03-29 修回日期:2023-09-18 出版日期:2024-05-15 发布日期:2024-05-08
  • 通讯作者: 曹建军(caojj@nudt.edu.cn)
  • 作者简介:(m15261820030@163.com)
  • 基金资助:
    国家自然科学基金(71901215,61371196);中国博士后科学基金(20090461425,201003797)

Specific Emitter Identification Based on Hybrid Feature Selection

GU Chumei1,2,3, CAO Jianjun1,2, WANG Baowei3, XU Yuxin1,2,3   

  1. 1 The Sixty-third Research Institute,National University of Defense Technology,Nanjing 210007,China
    2 Laboratory for Big Data and Decision,National University of Defense Technology,Changsha 410073,China
    3 School of Computer Science,Nanjing University of Information Science and Technology,Nanjing 210044,China
  • Received:2023-03-29 Revised:2023-09-18 Online:2024-05-15 Published:2024-05-08
  • About author:GU Chumei,born in 1997,postgra-duate,is a member of CCF(No.I7490G).Her main research interests include intelligent data analysis and application.
    CAO Jianjun,born in 1975,Ph.D,associate researcher,is a member of CCF(No.13414S).His main research intere- stsinclude data quality control and data intelligent analysis.
  • Supported by:
    National Natural Science Foundation of China(71901215,61371196) and China Postdoctoral Science Foundation Foundation(20090461425,201003797).

PDF (PC)

331

摘要： 为提高辐射源个体识别的准确率和运算效率,提出了一种基于混合式特征选择的辐射源个体识别。封装式特征选择方法分类正确率高,但计算复杂度高,处理高维数据时效率低。嵌入式特征选择方法计算复杂度低,但依赖于特定分类器。针对上述问题,综合封装式和嵌入式特征选择方法的特点,首先对信号数据使用3种嵌入式方法(随机森林、XGBoost和Ligh-tGBM)初选特征,分别得到随机森林子集、XGBoost子集和LightGBM子集。然后使用封装式方法对初选后得到的子集进行第二次降维,其中搜索策略分别使用序列后向搜索策略和蚁群优化算法,分类算法使用LightGBM。混合式方法共得到6种特征选择模型,通过对比各个模型得到的分类正确率和最优子集中的特征个数,确定最佳混合式特征选择模型。

关键词: 辐射源个体识别, 特征选择, 随机森林, XGBoost, LightGBM, 序列后向搜索策略, 蚁群优化

Abstract: To improve the accuracy and computational efficiency of specific emitter identification,a specific emitter identification based on hybrid feature selection is proposed.Wrapped feature selection methods have high classification accuracy,but it has high computational complexity and low efficiency in processing high-dimensional data.Embedded feature selection methods have low computational complexity,but rely on specific classifiers.To address the above problems,combining the characteristics of wrapped and embedded feature selection methods,firstly,three embedded methods(Random Forest,XGBoost,and LightGBM) are used to initially select features for signal data,and a random forest subset,an XGBoost subset and a LightGBM subset are obtained respectively.Secondly,the wrapped methods are used to perform a second dimensionality reduction on the subset obtained after the primary selection.Sequential backward selection and an ant colony optimization algorithm are used as research strategies respectively,while LightGBM is used as the classification algorithm.A total of six feature selection models are obtained from the proposed hybrid feature selection method.The optimal hybrid feature selection model is determined by comparing the classification accuracy and the number of features in the optimal subset obtained by each model.

Key words: Specific emitter identification, Feature selection, Random forest, XGBoost, LightGBM, Sequential backward selection, Ant colony optimization

中图分类号: 

  • TP391
[1]ZHANG M,LUO Z H,HUANG J G,et al.A fingerprint extraction method based on I/Q imbalance[J].Acta Electronica Sinica,2020,48(4):717-722.
[2]TANG Z,LEI Y K.Method of individual communication transmitter identification based on maximum correntropy[J].Journal on Communication,2016,37(12):1-5.
[3]LIU J F,YU H Y,DU J P,et al.Specific emitter identification under dynamic noise based on domain adaptation[J].Journal of Signal Processing,2021,37(6):1000-1007.
[4]DASGUPTA A,DRINEAS P,HARB B,et al.Feature selection methods for text classification[C]//Proceedings of the 13th ACM SIGKDD International Conference on Knowledge Disco-very and Data Mining.ACM,2007:230-239.
[5]HE T,HU J,XIA P,et al.Feature selection of Emg signal based on ReliefF algorithm and genetic algorithm[J].Journal of Shanghai Jiao Tong University,2016,50(2):204-208.
[6]WOZNICA A,NGUYEN P,KALOUSIS A.Model mining for robust feature selection[C]//Proceedings of the 18th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining.ACM,2012:913-921.
[7]TANG C,ZHENG X,LIU X W,et al.Cross-view locality preserved diversity and consensus learning for multi-view unsupervised feature selection[J].IEEE Transactions on Knowledge and Data Engineering,2022,34(10):4705-4716.
[8]JOHANNES H,MARTIN P,KLAUS B,et al.Leveraging mo-del inherent variable importance for stable online feature selection[C]//Proceedings of the 26th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining.ACM,2020:1478-1488.
[9]LI H,PUN C M,XU P,et al.A hybrid feature selection based on a discrete artificial beecolony for Parkinson's diagnosis[J].ACM Transactions on Internet Technology,2021,21(3):1-22.
[10]SHARMA D K,VARSHNEY R P,GARY A,et al.Hybrid feature selection method with multi-objective grey wolf optimizer for high dimension data[C]//Proceedings of the 9th Conference on Computing for Sustainable Global Development(INDIACom).IEEE,2022:854-859.
[11]LU X Q,HE W D,LU Q Q,et al.Hybrid filter-wrapper feature selection using water wave optimization for financial crisis prediction in enterprises[C]//Proceedings of the 16th Conference on Intelligent System and Knowledge(ISKE).IEEE,2021:193-199.
[12]ALYASIRI O M,CHEAH Y,ABASI A K.Hybrid filter-wrapper text feature selection technique for text classification[C]//Proceedings of the International Conference on Communication and Information Technology(ICICT).IEEE,2021:80-86.
[13]XU Z Z,SHEN D R,NIE T Z,et al.Hybrid features election algorithm combining information gain ratio and genetic algorithm[J].Journal of Software,2022,33(3):1128-1140.
[14]MCNAMARA Q,VEGA A D L,YARKONI T.Developing acomprehensive framework for multimodal feature extraction[C]//International Conference on Knowledge Discovery and Data Mining(SIGKDD).ACM,2017:1567-1574.
[15]CAO J J,ZHANG P L,ZHANG Y T,et al.Feature extraction of an engine cylinder head vibration signal based on lifting wavelet package transformation[J].Journal of Vibration and Shook,2008,27(2):34-37.
[16]XU Z L,JIN R,YE J P,et al.Non-monotonic feature selection[C]//Proceedings of the 26th Annual International Conference on Machine Learning.ACM,2009:1145-1152.
[17]JUN Y J,LEAU Y,ALIAS S,et al.A multi-filter feature selection in detecting distributed denial-of-service attack[C]//Proceedings of the 3rd International Conference on Telecommunications and Communication Engineering.ACM,2019:57-62.
[18]FAN X,FENG Z Q,YANG X H,et al.Hazw weather recognition based on multiple features and random forest[C]//Procee-dings of International Conference on Security,Pattern Analysis,and Cybernetics.IEEE,2018:485-488.
[19]FENG D D,DENG Z F,WANG T X,et al.Identification ofdisturbance sources based on random forest model[C]//Proceedings of International Conference on Power System Techno-logy.IEEE,2018:3370-3375.
[20]ZHAI Y B,ZHENG X H.Random forest traffic classification method in SDN[C]//Proceedings of International Conference on Cloud Computing,Big Data and Blockchain.IEEE,2018:1-5.
[21]CHEN T,GUESTRIN C.XGBoost:A scalable tree boosting system[C]//Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining.ACM,2016:785-794.
[22]LI L,LI C,WU Y,et al.Spectroscopy-based food internal quality evaluation with XGBoost algorithm[C]//APWeb-WAIM 2018.Springer,2018:56-64.
[23]LI Z S,YAO X,LIU Z G,et al.Feature selection algorithmbased on LightGBM[J].Journal of Northeastern University(Natural Science),2021,42(12):1688-1695.
[24]YUAN J,RAO Z,LIN H,et al.Classification of Chinese dialect regions from L2 English speech[C]//International Conference on Acoustics,Speech and SP(ICASSP).IEEE,2019:8117-8121.
[25]SHEN S,QIAN Y,ZHENG J,et al.Accurately predicting circ-RNA-disease associations using variational graph auto-encoders and LightGBM[C]//International Conference on Bioinformatics and Biomedicine(BIBM).Piscataway,IEEE,2021:522-527.
[26]CAO J J,ZHANG P L,WANG Y X,et al.Graph-based ant system for subset problems[J].Journal of System Simulation,2008,20(22):6146-6150.
[27]CHEN Z,LV N.Network intrusion detection model based onRandom Forest andXGBoost[J].Journal of Signal Processing,2020,36(7):1055-1064.
[28]CAO J J,GU C M,WANG B W,et al.Specific emitter identification based on ACO-XGBoost [C]//APWeb-WAIM 2022.Springer,2023:76-90.
[29]GU C M,CAO J J,WANG B W,et al.Specific emitter identification of LightGBM based on colony parameters optimization[J].Journal of Computer Engineering and Science,2023,45(1):9-18.
Viewed
Full text


Abstract

Cited
  Shared   
  Discussed   
No Suggested Reading articles found!
渝ICP备16013121号-4
地址：重庆市渝北区洪湖西路18号    邮编：401121  
电话：023-63500828（编辑部） 023-67039612（会议/专题） 023-67039623（财务/发票）
E-mail：jsjkx12@163.com
本系统由北京玛格泰克科技发展有限公司设计开发