• 
    

    
    

      99热精品在线国产_美女午夜性视频免费_国产精品国产高清国产av_av欧美777_自拍偷自拍亚洲精品老妇_亚洲熟女精品中文字幕_www日本黄色视频网_国产精品野战在线观看 ?

      基于參數(shù)優(yōu)化VMD和LightGBM的雷達(dá)輻射源個體識別

      2022-05-18 10:47:01肖易寒李棟年于祥禎宋柯
      航空兵器 2022年2期

      肖易寒 李棟年 于祥禎 宋柯

      摘 要:????? 為解決在復(fù)雜電磁環(huán)境中雷達(dá)輻射源個體識別準(zhǔn)確率低的問題,提出基于參數(shù)優(yōu)化VMD和LightGBM的雷達(dá)輻射源個體識別技術(shù)。首先對雷達(dá)輻射源的無意特征進(jìn)行分析,仿真添加了相位噪聲作為雷達(dá)輻射源的指紋特征; 其次利用麻雀搜索算法(SSA)對變分模態(tài)分解(VMD)的分解參數(shù)進(jìn)行自動尋優(yōu),準(zhǔn)確快速地得到最優(yōu)分解參數(shù)組合為[2, 2 950]; 然后基于最優(yōu)VMD分解參數(shù)對輻射源信號提取能量熵與樣本熵作為特征向量; 最后將特征向量送入LightGBM分類器完成輻射源個體識別。通過實測數(shù)據(jù)的驗證,信噪比在25 dB時識別率能夠達(dá)到85%以上,具有較為理想的識別結(jié)果。

      關(guān)鍵詞:???? 雷達(dá)輻射源; 個體識別; 變分模態(tài)分解; 麻雀搜索算法; 能量熵; 樣本熵; LightGBM

      中圖分類號:???? TJ760; V243.2

      文獻(xiàn)標(biāo)識碼:??? A

      文章編號:???? 1673-5048(2022)02-0093-08

      DOI: 10.12132/ISSN.1673-5048.2021.0073

      0 引? 言

      輻射源的個體識別又稱輻射源指紋識別或者特定輻射源識別[1],輻射源指紋是發(fā)射設(shè)備硬件固有非理想特征造成的,具有不可偽造、難以改變、不可避免等特點。這種非理想特征對信號的影響是細(xì)微的,以無意調(diào)制的形式附加在發(fā)射信號上,對“指紋”特征的提取是輻射源無意識別的難點。目前對指紋特征的分析提取主要基于對以下信息的獲取: 常規(guī)基本參數(shù)信息[2-4]、基本變換信息[5-8]、信號特殊結(jié)構(gòu)[9-11]、分解重構(gòu)信息[12-15]以及發(fā)射機(jī)硬件特性[16-18]。

      輻射源個體識別可根據(jù)分析對象分為通信輻射源個體識別與雷達(dá)輻射源個體識別[19]。其中對于雷達(dá)輻射源個體識別的研究具有重要意義,快速精確地對雷達(dá)輻射源進(jìn)行識別可以直接掌握戰(zhàn)爭主動性,是當(dāng)今電子戰(zhàn)快速發(fā)展下的制勝關(guān)鍵。

      雷達(dá)輻射源個體識別并不關(guān)注傳輸過程信號的主要信息,關(guān)注重點是在信號主體成分之外的細(xì)微特征。利用信號的分解算法可以在提高信息維度的同時獲得新的數(shù)據(jù)處理思路,以此得到更好的特征提取結(jié)果。分解算法可以將信號分成多個本征模態(tài)函數(shù)(IMF),分別為原始信號的主要成分及雜散成分。常見的分解算法有經(jīng)驗?zāi)B(tài)分解(EMD)、變分模態(tài)分解(VMD)及固有時間尺度分解(ITD),其中VMD方法結(jié)果穩(wěn)定,計算簡單,無模態(tài)混疊問題,分解出的基本分量IMF具有AM-FM調(diào)制窄帶信號的特點,其瞬時頻率有實際的物理意義。馬洪斌等[20]利用蛙跳算法來搜索VMD最優(yōu)參數(shù),將得到的IMF分量構(gòu)成矩陣進(jìn)行奇異值分解,并作為特征對故障類型進(jìn)行識別。李亞蘭等[21]將VMD算法用于雷達(dá)信號有意識別,將信號分解為3個IMF分量后對3個IMF分量提取排列熵與樣本熵,在低信噪比下達(dá)到較高識別率。鄭義等[22]提出利用蝗蟲優(yōu)化算法,以相關(guān)峭度為適應(yīng)度函數(shù)對變分模態(tài)分解參數(shù)進(jìn)行自適應(yīng)選定,用于提取強(qiáng)噪聲背景下滾動軸承故障信號的微弱特征。

      本文將麻雀搜索算法(SSA)用于VMD參數(shù)的優(yōu)化選取上,解決了VMD分解參數(shù)人為設(shè)置帶來的影響,又基于熵特征進(jìn)行了二次特征提取,最后用于LightGBM分類器進(jìn)行識別。 仿真結(jié)果顯示該方法具有較高識別率,較其他方法有明顯優(yōu)勢,在實測數(shù)據(jù)的驗證下同樣能夠達(dá)到一定識別率。

      3.3 基于LightGBM的輻射源個體識別

      仿真得到的雷達(dá)輻射源信號經(jīng)過SSA優(yōu)化VMD參數(shù)后分解得到3個IMF分量,分別提取了能量熵與樣本熵后串聯(lián)為6維特征向量,使用LightGBM進(jìn)行分類識別。每個信噪比下有3部輻射源共1 200個樣本信號,分類識別中將訓(xùn)練集與測試集比值設(shè)置為8∶2,即960個樣本作為訓(xùn)練集,240個樣本為測試集,送入LightGBM進(jìn)行分類識別。

      將本文方法與文獻(xiàn)[31]中的方法進(jìn)行對比。文獻(xiàn)[31]對信號進(jìn)行VMD后將模態(tài)分量不用的中心頻率作為特征,送入經(jīng)人工蜂群算法優(yōu)化的支持向量機(jī)(ABC-SVM)中進(jìn)行識別分類。同時,對比了使用本文方法提取特征后送入ABC-SVM的識別率以及文獻(xiàn)[31]所提取特征送入LightGBM分類器的識別率,如圖6所示。

      由圖6可得,識別率隨信噪比的增加逐步提升,本文方法即VMD分解后提取能量熵與樣本熵作為特征向量后使用LightGBM分類器進(jìn)行分類識別,在15 dB時3部輻射源識別率已經(jīng)達(dá)到了95%以上,在20 dB時已經(jīng)達(dá)到了100%。

      對比分析可得:

      (1) 本文識別結(jié)果較文獻(xiàn)[31]方法結(jié)果在各個信噪比下識別率均有所提升,尤其在較低信噪比下的識別率提升較高。

      (2) 本文所提取特征經(jīng)ABC-SVM識別方法在各個信噪比下識別率均高于文獻(xiàn)[31]方法,這表明在使用相同分類器的情況下,本文所提取特征對輻射源無意特征的表征要優(yōu)于文獻(xiàn)[31]所提取特征。

      (3) 對文獻(xiàn)[31]所提取特征經(jīng)LightGBM識別方法后在各個信噪比下識別率均略高于文獻(xiàn)[31]方法,這表明在使用相同特征的情況下,本文所選用分類器同樣優(yōu)于ABC-SVM。

      (4) 以文獻(xiàn)[31]方法為參考對象,本文所提取特征對識別率的提升效果要優(yōu)于選用LightGBM對識別率的提升效果。

      3.4 基于實測數(shù)據(jù)的識別結(jié)果分析

      為了驗證實驗方法的穩(wěn)定性以及對輻射源信號中添加無意特征的仿真有效性,采集了來自3部不同型號、仿真參數(shù)相同的信號源發(fā)射器發(fā)出的線性調(diào)頻信號。

      實驗采用3臺信號發(fā)射器,分別為Agilent E4438C,Agilent N5172B EXG X以及Tektronix AWG70001。對每部信號發(fā)生器分別采集400組樣本數(shù)據(jù),其中80%用于訓(xùn)練,剩余20%用于測試。

      對采集的信號進(jìn)行處理后,經(jīng)基于SSA的VMD參數(shù)優(yōu)化算法進(jìn)行分析,得到最優(yōu)分解參數(shù)組合為[2, 2 950],即分解模態(tài)數(shù)為2,二次懲罰因子為2 950。后經(jīng)VMD分解后得到2個模態(tài)分量,對其分別提取能量熵與樣本熵后級聯(lián)得4維特征向量,送入LightGBM得到的分類結(jié)果如表2所示。

      由實驗結(jié)果可以看出,識別率隨信噪比的增大逐漸提升。信噪比在25 dB時識別率能夠達(dá)到85%以上,在雷達(dá)輻射源個體識別研究問題中能夠達(dá)到較高識別率。4 結(jié)? 論

      (1) VMD算法能夠依據(jù)不同中心頻率對信號進(jìn)行分解,在輻射源無意特征的提取方面有天然的優(yōu)勢。同時SSA又能夠快速準(zhǔn)確地對VMD的分解模態(tài)數(shù)K與二次懲罰因子α進(jìn)行尋優(yōu),避免了人為選擇參數(shù)對分解結(jié)果的不利影響。能量熵與樣本熵完全匹配于VMD算法,將各模態(tài)分量間的重要信息量化表達(dá)。仿真結(jié)果顯示,基于LightGBM算法的識別結(jié)果要好于經(jīng)典的SVM,本文所提方法各方面性能均有穩(wěn)定提升。

      (2) 經(jīng)實測數(shù)據(jù)驗證,本文方法能夠達(dá)到較高識別率,但相較于仿真數(shù)據(jù)識別率有一定差距,證明雷達(dá)輻射源中的相位噪聲是雷達(dá)發(fā)射機(jī)中無意特征的主要來源,但仿真的相位噪聲距離實際無意特征還有一定差距,還需進(jìn)一步仿真分析,以求更加接近真實的無意特征。

      參考文獻(xiàn):

      [1] Ren D F, Zhang T. Specific Emitter Identification Based on Intrinsic Time-Scale-Decomposition and Image Texture Feature[C]∥9th IEEE International Conference on Communication Software and Networks (ICCSN), 2017: 1302-1307.

      [2] Wu L W, Zhao Y Q, Feng M F, et al. Specific Emitter Identification Using IMF-DNA with a Joint Feature Selection Algorithm[J]. Electronics, 2019, 8(9): 934.

      [3] Feng Z P, Zhang D, Zuo M J. Adaptive Mode Decomposition Me-thods and Their Applications in Signal Analysis for Machinery Fault Diagnosis: A Review with Examples[J]. IEEE Access, 2017, 5: 24301-24331.

      [4] Ru X H, Liu Z, Huang Z T, et al. Evaluation of Unintentional Modulation for Pulse Compression Signals Based on Spectrum Asymmetry[J]. IET Radar, Sonar & Navigation, 2017, 11(4): 656-663.

      [5] 葉文強(qiáng), 俞志富. 基于聯(lián)合時頻輻射源信號識別方法[J]. 電子信息對抗技術(shù), 2018, 33(5): 16-19.

      Ye Wenqiang, Yu Zhifu. Signal Recognition Method Based on Joint Time-Frequency Radiant Source[J]. Electronic Information Warfare Technology, 2018, 33(5): 16-19.(in Chinese)

      [6] Flamant J, Le Bihan N, Chainais P. Time-Frequency Analysis of Bivariate Signals[J]. Applied and Computational Harmonic Analysis, 2019, 46(2): 351-383.

      [7] 王歡歡, 張濤. 結(jié)合時域分析和改進(jìn)雙譜的通信信號特征提取算法[J]. 信號處理, 2017, 33(6): 864-871.

      Wang Huanhuan, Zhang Tao. Extraction Algorithm of Communication Signal Characteristics Based on Improved Bispectra and Time-Domain Analysis[J]. Journal of Signal Processing, 2017, 33(6): 864-871.(in Chinese)

      [8] Zhang J W, Wang F G, Dobre O A, et al. Specific Emitter Identification via Hilbert-Huang Transform in Single-Hop and Relaying Scenarios[J]. IEEE Transactions on Information Forensics and Security, 2016, 11(6): 1192-1205.

      [9] Padilla P, Padilla J L, Valenzuela-Valdés J F. Radio Frequency Identification of Wireless Devices Based on RF Fingerprinting[J]. Electronics Letters, 2013, 49(22): 1409-1410.

      [10] Wu Q Y, Feres C, Kuzmenko D, et al. Deep Learning Based RF Fingerprinting for Device Identification and Wireless Security[J]. Electronics Letters, 2018, 54(24): 1405-1407.

      [11] 王磊, 姬紅兵, 史亞. 基于模糊函數(shù)代表性切片的運動雷達(dá)輻射源識別[J]. 系統(tǒng)工程與電子技術(shù), 2010, 32(8): 1630-1634.

      Wang Lei, Ji Hongbing, Shi Ya. Moving Radar Emitter Recognition Based on Representative-Cut Feature of Ambiguity Function[J]. Systems Engineering and Electronics, 2010, 32(8): 1630-1634.(in Chinese)

      [12] Martis R J, Acharya U R, Tan J H, et al. Application of Intrinsic Time-Scale Decomposition (ITD) to EEG Signals for Automated Seizure Prediction[J]. International Journal of Neural Systems, 2013, 23(5): 1350023.

      [13] Xu Y G, Xie Z C, Cui L L, et al. The Feature Extraction Method of Gear Magnetic Memory Signal[J]. Advanced Materials Research, 2013, 819: 206-211.

      [14] Bihl T J, Bauer K W, Temple M A. Feature Selection for RF Fingerprinting with Multiple Discriminant Analysis and Using Zig Bee Device Emissions[J]. IEEE Transactions on Information Forensics and Security, 2016, 11(8): 1862-1874.

      [15] Zhu S L, Gan L. Specific Emitter Identification Based on Visibility Graph Entropy[J]. Chinese Physics Letters, 2018, 35(3): 030501.

      [16] 徐志軍, 陳志偉, 王金明, 等. 基于功放特性的輻射源識別的改進(jìn)方法[J]. 南京郵電大學(xué)學(xué)報:自然科學(xué)版, 2013, 33(6): 54-58.

      Xu Zhijun, Chen Zhiwei, Wang Jinming, et al. An Improved Method for Emitter Identification Based on Character of Power Amplifier[J]. Journal of Nanjing University of Posts and Telecommunications: Natural Science, 2013, 33(6): 54-58.(in Chinese)

      [17] 黃淵凌, 鄭輝. 一種基于相噪特性的輻射源指紋特征提取方法[J]. 計算機(jī)仿真, 2013, 30(9): 182-185.

      Huang Yuanling, Zheng Hui. Emitter Fingerprint Feature Extraction Method Based on Characteristics of Phase Noise[J]. Computer Simulation, 2013, 30(9): 182-185.(in Chinese)

      [18] Wu L W, Zhao Y Q, Wang Z, et al. Specific Emitter Identification Using Fractal Features Based on Box-Counting Dimension and Variance Dimension[C]∥IEEE International Symposium on Signal Processing and Information Technology (ISSPIT), 2017: 226-231.

      [19] Talbot K I, Duley P R, Hyatt M H. Specific Emitter Identification and Verification[J]. Technology Review Journal, 2003,1(1): 113-133.

      [20] 馬洪斌, 佟慶彬, 張亞男. 優(yōu)化參數(shù)的變分模態(tài)分解在滾動軸承故障診斷中的應(yīng)用[J]. 中國機(jī)械工程, 2018, 29(4): 390-397.

      Ma Hongbin, Tong Qingbin, Zhang Yanan. Applications of Optimization Parameters VMD to Fault Diagnosis of Rolling Bearings[J]. China Mechanical Engineering, 2018, 29(4): 390-397.(in Chinese)

      [21] 李亞蘭, 金煒東, 葛鵬. 基于VMD和特征融合的輻射源信號識別[J]. 系統(tǒng)工程與電子技術(shù), 2020, 42(7): 1499-1503.

      Li Yalan, Jin Weidong, Ge Peng. Radiation Emitter Signal Reco-gnition Based on VMD and Feature Fusion[J]. Systems Engineering and Electronics, 2020, 42(7): 1499-1503.(in Chinese)

      [22] 鄭義, 岳建海, 焦靜, 等. 基于參數(shù)優(yōu)化變分模態(tài)分解的滾動軸承故障特征提取方法[J]. 振動與沖擊, 2021, 40(1): 86-94.

      Zheng Yi, Yue Jianhai, Jiao Jing, et al. Fault Feature Extraction Method of Rolling Bearing Based on Parameter Optimized VMD[J]. Journal of Vibration and Shock, 2021, 40(1): 86-94.(in Chinese)

      [23] Dragomiretskiy K, Zosso D. Variational Mode Decomposition[J]. IEEE Transactions on Signal Processing, 2014, 62(3): 531-544.

      [24] Xue J K, Shen B. A Novel Swarm Intelligence Optimization Approach: Sparrow Search Algorithm[J]. Systems Science & Control Engineering, 2020, 8(1): 22-34.

      [25] Alcaraz R, Rieta J J. A Review on Sample Entropy Applications for the Non-Invasive Analysis of Atrial Fibrillation Electrocardiograms[J]. Biomedical Signal Processing and Control, 2010, 5(1): 1-14.

      [26] Pincus S M. Approximate Entropy as a Measure of Irregularity for Psychiatric Serial Metrics[J]. Bipolar Disorders, 2006, 8(5 Pt 1): 430-440.

      [27] Ke G L, Meng Q, Finley T, et al. LightGBM: A Highly Efficient Gradient Boosting Decision Tree [C]∥ 31st International Conference on Neural Information Processing Systems, 2017: 3149-3157.

      [28] 鄭娜娥, 王盛, 張靖志, 等. 基于射頻指紋的輻射源個體識別技術(shù)綜述[J]. 信息工程大學(xué)學(xué)報, 2020, 21(3): 285-289.

      Zheng Na’e,Wang Sheng, Zhang Jingzhi, et al. Overview of Emitter Identification Techniques Based on Radio Frequency Fingerprinting[J]. Journal of Information Engineering University, 2020, 21(3): 285-289.(in Chinese)

      [29] 孫麗婷, 黃知濤, 王翔, 等. 輻射源指紋特征提取方法述評[J]. 雷達(dá)學(xué)報, 2020, 9(6): 1014-1031.

      Sun Liting, Huang Zhitao, Wang Xiang, et al. Overview of Radio Frequency Fingerprint Extraction in Specific Emitter Identification[J]. Journal of Radars, 2020, 9(6): 1014-1031.(in Chinese)

      [30] 楊冬鋒, 陳盛開, 劉曉軍, 等. 基于自適應(yīng)VMD和時-頻分段能量熵特征的過電壓信號識別[J]. 電網(wǎng)技術(shù), 2019, 43(12): 4597-4604.

      Yang Dongfeng, Chen Shengkai, Liu Xiaojun, et al. Research on Overvoltage Signal Recognition Based on Adaptive VMD and Time-Frequency Segment Energy Entropy[J]. Power System Technology, 2019, 43(12): 4597-4604.(in Chinese)

      [31] 張忠民, 劉剛. 基于VMD和ABC-SVM的雷達(dá)輻射源個體識別[J]. 哈爾濱商業(yè)大學(xué)學(xué)報:自然科學(xué)版, 2020, 36(2): 176-182.

      Zhang Zhongmin, Liu Gang. Individual Identification of Radar Emitter Based on VMD and ABC-SVM[J]. Journal of Harbin University of Commerce: Natural Sciences Edition, 2020, 36(2): 176-182.(in Chinese)

      Radar Emitter Individual Identification Based on

      Parameter Optimization VMD and LightGBM

      Xiao Yihan1,Li Dongnian1*,Yu Xiangzhen2,Song Ke2

      (1. Harbin Engineering University,Harbin 150001, China;

      2. Shanghai Radio Equipment Research Institute,Shanghai 200090, China)

      Abstract: In order to solve the problem of low accuracy of? radar emitter individual identification in? complex electromagnetic environment, a radar emitter individual identification technology based on parameter optimization VMD and LightGBM is proposed. Firstly, the unintentional features of the radar emitter are analyzed, and the added phase noise is taken as the fingerprint feature of? radar emitter in the simulation. Secondly,sparrow search algorithm (SSA) is used to automatically optimize the decomposition parameters of? variational modal decomposition (VMD), and the optimal decomposition parameter combination is accurately and quickly obtained as [2, 2 950]. Then, based on the optimal VMD decomposition parameters, the energy entropy and sample entropy of the emitter signal are extracted as? feature vector. Finally, the feature vector is sent to the LightGBM classifier to complete the emitter individual identification. Through the verification of measured data,the recognition rate can reach more than 85% when the signal-to-noise ratio is 25 dB, which has? ideal recognition results.

      Key words:? radar emitter; individual identification; VMD; SSA; energy entropy; sample entropy; LightGBM

      河南省| 榆树市| 苍溪县| 镇原县| 本溪| 高平市| 亳州市| 顺义区| 德惠市| 沐川县| 平利县| 吉林市| 南宫市| 赤峰市| 长海县| 建平县| 仁寿县| 汝南县| 疏勒县| 关岭| 镇坪县| 潍坊市| 乐平市| 龙口市| 高邮市| 藁城市| 伊吾县| 彝良县| 南华县| 衡阳县| 黔江区| 繁昌县| 航空| 长子县| 长治市| 砀山县| 房产| 光山县| 南陵县| 靖安县| 罗定市|