基于多域特征融合的通信辐射源个体识别方法

doi:10.12382/bgxb.2021.0880

摘要/Abstract

摘要：

为解决利用单一特征进行通信辐射源个体识别识别率不高的问题,提出一种基于多域特征融合的通信辐射源个体识别方法。提取通信辐射源发射信号的多个变换域特征,并组合这些特征为多域特征。构建多通道卷积神经网络,利用多通道卷积操作对多域特征进行深层次提取。通过神经网络的分类器,完成对通信辐射源个体的分类。在低信噪比和瑞利信道条件下,使用所提方法对20个CC2530设备进行识别。研究结果表明:与基于单一特征的辐射源个体识别方法相比,该方法充分利用了通信辐射源发射信号的多个变换域特征,结合神经网络的强大细微特征挖掘能力,实现了对通信辐射源个体的有效识别;该方法能够显著提升在低信噪比的识别准确率和时效性,在0dB条件下的识别效果仍可达到91.01%。

关键词: 通信辐射源识别, 特征融合, 多通道卷积神经网络

Abstract:

To solve the problem of low identification rate caused by specific communication emitter identification using a single feature, a method for individual identification of the communication emitter based on multi-domain feature fusion is proposed. Firstly, multiple transform domain features of the signals transmitted by the communication emitter are extracted, and these features are combined into multi-domain features.Secondly, the multi-channel convolution neural network is constructed, and the multi-channel convolution operation is carried out to extract the multi-domain features at a deep level. Finally, specificcommunication emitter classification is completed using the neural network classifier.Compared with the identification method based on a single feature, this method makes full use of the multi-domain features of the signals sent by the communication emitter and combines the powerful microscopic feature mining capability of the neural network to realize the effective individual identification of the communication emitter. Through the identification of 20 CC2530 devices under the conditions of low signal-to-noise ratio(SNR) and Rayleigh channel, the results show that the proposed method can significantly improve the identification accuracy and timeliness under a low SNR, and that the identification effect can still reach 91.01% under the condition of 0dB.

Key words: communication emitter identification, feature fusion, multi-channel convolution neural network

王检, 张邦宁, 张洁, 魏国峰, 郭道省. 基于多域特征融合的通信辐射源个体识别方法[J]. 兵工学报, 2023, 44(4): 949-959.

WANG Jian, ZHANG Bangning, ZHANG Jie, WEI Guofeng, GUO Daoxing. A Method for Specific Communication Emitter Identification Based on Multi-Domain Feature Fusion[J]. Acta Armamentarii, 2023, 44(4): 949-959.

图/表 14

图1 基于多域特征的辐射源个体识别模型

Fig.1 Specific communication emitter identification model based on multi-domain features

图2 辐射源及数据采集系统

Fig.2 Emitter and data acquisition system

图3 标准化的前同步码I/Q数据

Fig.3 Standardized preamble I/Q data

图4 Welch算法的分段方法

Fig.4 Segmentation method of the Welch algorithm

图5 0dB信噪比前同步码的功率谱及频谱

Fig.5 Power spectrum and spectrum of preamble with SNR of 0dB

图6 前同步码双谱等高线图

Fig.6 Bispectrum contour map of preamble

图7 矩形积分路径

Fig.7 Rectangle integration path

图8 特征融合数据示意图

Fig.8 Schematic diagram of feature fusion data

图9 多通道一维卷积神经网络结构模型

Fig.9 Model of multi-channel one-dimensional CNN structure

图10 瑞利信道条件下的I/Q数据

Fig.10 I/Q data in Rayleigh Channel

表1 不同训练轮次对0dB信号的识别结果

Table 1 Identification results of different training epochs on 0dB signal%

实验方法	epoch
实验方法	4	8	12	16	20
多域特征融合	87.08	89.10	89.92	90.45	91.01
基于I/Q数据	66.74	75.01	80.78	84.01	85.17
基于功率谱数据	71.76	79.70	84.15	86.63	87.55
基于积分双谱数据	61.50	68.89	72.02	74.11	76.50

图11 单个变换域与多域特征融合的识别效果对比

Fig.11 Comparison of identification effect between single transform domain and multi-domain feature fusion

表2 20轮次单个变换域与多域特征融合的识别结果

Table 2 Identification results of single transform domain and multi-domain feature fusion at 20 epochs%

实验方法	信噪比/dB
实验方法	0	5	10	15	20	25
多域特征融合	91.01	94.83	96.58	98.15	98.48	98.98
基于I/Q数据	85.17	89.23	90.75	92.11	93.25	93.88
基于功率谱数据	87.55	91.93	93.25	93.35	93.60	94.00
基于积分双谱数据	76.50	83.30	88.50	91.15	92.55	93.05

图12 不同算法的识别效果对比

Fig.12 Comparison of identification effects of different algorithms

参考文献 33

[1]	XU Q, ZHENG R, SAAD W, et al. Device fingerprinting in wireless networks:challenges and opportunities[J]. IEEE Communications Surveys & Tutorials, 2016, 18(1):94-104.
[2]	WANG S, PENG L N, FU H, et al. Aconvolutional neural network-based RF fingerprinting identification scheme for mobile phones[C]// Proceedings of IEEE INFOCOM 2020-IEEE Conference on Computer Communications Workshops.Toronto, ON, Canada: IEEE Press, 2020:15-120.
[3]	PENG L N, ZHANG J Q, LIU M, et al. Deep learning based RF fingerprint identification using differential constellation trace figure[J]. IEEE Transactions on Vehicular Technology, 2019, 69(1):1091-1095. doi: 10.1109/TVT.25 URL
[4]	TAMURA H, ANAGISAWA K, SHIRANE A, et al. Wireless devices identification with light-weight convolutional neural network operating on quadrant IQ transition image [C]//Proceedings of the 2020 18th IEEE International New Circuits and Systems Conference. Montreal, QC, Canada: IEEE Press, 2020:106-109.
[5]	RU X H, LIU Z, HUANG Z T, et al. Evaluation of unintentional modulation for pulse compress-ion signals based on spectrum asymmetry[J]. IET Radar, Sonar & Navigation, 2017, 11(4):656-663. doi: 10.1049/rsn2.v11.4 URL
[6]	WANG W, LIU H, YANG J A, et al. Specific emitter identification using decomposed hierarchical feature extraction methods[C]// Proceedings of International Conference on Natural Computation.Piscataway, NJ, US: IEEE Press, 2017:1639-1643.
[7]	WONG L J, HEADLEY W C, MICHAELS A J. Specific emitter identification using convolutional neural network-based IQ imbalance estimators[J]. IEEE Access, 2019, 7:33544-33555. doi: 10.1109/Access.6287639 URL
[8]	RAM M, BANER C, ROY D, et al. Exploitingspatio-temporal properties of I/Q signal data using 3D convolution for RF transmitter ientification[J]. IEEE Journal of Radio Frequency Identification, 2021, 99:113-127.
[9]	LING Y, JIA J C, WANG S, et al. Wireless device identification based on radio frequency fingerprint features[C]//Proceedings of the 2020 IEEE International Conference on Communications. Dublin, Ireland:IEEE, 2020:1-6.
[10]	COGHILL C, REHMAN S U, SOWERBY K W. Radio-frequency fingerprinting for mitigating primary user emulation attack in low-end cognitive radios[J]. IET Communications, 2014, 8(8):1274-128 doi: 10.1049/cmu2.v8.8 URL
[11]	DING L D, WANG S L, WANG F G, et al. Specific emitter identification via convolutional neural networks[J]. IEEE Communications Letters, 2018, 22(12):2591-2594. doi: 10.1109/LCOMM.2018.2871465 URL
[12]	YAO Y Y, YU L, CHEN Y M. Specific emitter identification based on square integral bispectrum features[C]//Proceedings of the 2020 IEEE 20th International Conference on Communication Technology. Nanning, China:IEEE, 2020:1311-1314.
[13]	GONG J L, XU X D, LEI Y K. Unsupervised specific emitter identification method using radio-frequency fingerprint embedded infoGAN[J]. IEEE Transactions on Information Forensics and Security, 2020, 15: 2898-2913. doi: 10.1109/TIFS.10206 URL
[14]	CHEN S L, XIE F Y, CHEN Y, et al. Identification of wireless transceiver devices using radio frequency (RF) fingerprinting based on STFT analysis to enhance authentication security[C]//Proceedings of the 2017 IEEE 5th International Symposium on Electromagnetic Compatibility. Beijing, China: IEEE, 2017:1-5.
[15]	TU Y, ZHANG Z, LI Y B, et al. Research on the Internet of Things device recognition based on RF-fingerprinting[J]. IEEE Access, 2019, 7:37426-37431. doi: 10.1109/Access.6287639 URL
[16]	ZHOU X Y, HU A Q, LI G Y, et al. A robust radio-frequency fingerprint extraction scheme for practical device recognition[J]. IEEE Internet of Things Journal, 2021, 99: 1276-11289.
[17]	Wireless medium access control (MAC) and physical layer (PHY) specifications for low-rate wireless personal area networks (WPANs):IEEE 802.15.4[S]. New York, NY, US:IEEE, 2006.
[18]	MOSTAFA A E, WONG V W, ZHOU Y, et al. Aggregate preamble sequence design and detection for massive IoT with deep learning[J]. IEEE Transactions on Vehicular Technology, 2021, 70(4):3800-3816. doi: 10.1109/TVT.2021.3064868 URL
[19]	RAMSEY B W, TEMPLE M A, MULLINS B E. PHY foundation for multi-factor ZigBee node authentication[C]//Proceedings of the 2012 IEEE Global Communications Conference. Anaheim, CA, US: IEEE Press, 2012:795-800.
[20]	Al-SHAWABKA A, RESTUCCIA F, D’ORO S, et al. Massive-scale I/Q datasets for WiFi radio fingerprinting[J]. Computer Networks, 2020, 182: 107566. doi: 10.1016/j.comnet.2020.107566 URL
[21]	MITRA S K. Digital signal processing:a computer-based approach[M]. 3rd ed. Tsinghua University Press, 2012.
[22]	PARHI K, AYINAL M. Low-complexity welch power spectral density computation[J]. IEEE Transactions on Circuits and Systems I:Regular Papers, 2014, 61(1):172-182. doi: 10.1109/TCSI.2013.2264711 URL
[23]	ZHANG G C, DU R X, SHI T L. Extracting gear fault features using integrated bispectrum[C]//Robotics Intelligent Systems and Signal Processing. Changsha, China: IEEE, 2003:548-553.
[24]	WEI D X, ZHANG S N, CHEN S, et al. Research on anti-jamming technology of chaotic composite shortrange detection system based on underdetermined signal separation and spectral analysis[J]. IEEE Access, 2019, 7:42298-42308. doi: 10.1109/Access.6287639 URL
[25]	FAILLI M. Digital land mobile radio communications:COST 207[R]. London: Commission of the European Communities, 1989.
[26]	FADUL M, REISING D, LOVELESS T, et al. RF-DNA fingerprint classification of OFDM signals using a rayleigh fading channel model[C]// Proceedings of the 2019 IEEE Wireless Communications and Networking Conference. Marrakesh, Morocco: IEEE Press, 2019:1-7.
[27]	李晨明. 非合作通信中的盲均衡技术研究[D]. 西安: 西安电子科技大学, 2020:111.
	LI C. Research on Equalization Technology in Non-cooperative Communication[D]. Xi’an: Xidian University, 2020:111. (in Chinese)
[28]	Al-SHAWABKA A, RESTUCCIA F, D’ORO S, et al. Exposing the fingerprint: dissecting the impact of the wireless channel on radio fingerprinting[C]//Proceedings of the IEEE INFOCOM 2020-IEEE Conference on Computer Communications. Toronto, Canada: IEEE, 2020:646-655.
[29]	石荣, 吴聪. 信噪比的不同定义及对调制识别性能评估的影响[J]. 通信技术, 2019, 52(7):1556-1562.
	SHI R, WU C. Different definitions of SNR and its influence on performance evaluation of modulation recognition[J]. Communications Technology, 2019, 52(7):1556-1562. (in Chinese)
[30]	PAN Y W, YANG S H, PENG H, et al. Specific emitter identification based on deep residual networks[J]. IEEE Access, 2019, 7:54425-54434. doi: 10.1109/ACCESS.2019.2913759
[31]	陈悦, 雷迎科, 李昕, 等. 基于IQ图特征的通信辐射源个体识别[J]. 信号处理, 2021, 37(1):120-125.
	CHEN Y, LEI Y K, LI X, et al. Specific emitter identification of communication radiation source based on the characteristics IQ graph features[J]. Journal of Signal Processing, 2021, 37(1):120-125. (in Chinese)
[32]	梁华东, 韩江洪. 采用双谱多类小波包特征的雷达信号聚类分选[J]. 光子学报, 2014, 43(3):158-165.
	LIANG H D, HAN J H. Clustering and sorting radar signals based on multi wavelet packets characteristics of bispectrum[J]. Acta Photonica Sinica, 2014, 43(3):158-165. (in Chinese)
[33]	ZHAO W Z, DU S H. Spectral-spatial feature extraction for hyperspectral image classifi-cation: a dimension reduction and deep learning approach[J]. IEEE Transactions on Geoscience and Remote Sensing, 2016, 54(8):4544-4554. doi: 10.1109/TGRS.2016.2543748 URL