| 663 | 9 | 517 |
| 下载次数 | 被引频次 | 阅读次数 |
随着第五代移动通信、物联网等技术的普及，辐射源数量与类型都呈现井喷式增长，这导致基于统计特征与机器学习的传统特定辐射源识别技术的识别性能难以达到实际应用水平。近年来，深度学习在计算机视觉、自然语言处理等领域表现出的高超性能启发了大量研究者将深度学习用于特定辐射源识别问题，并且取得了丰硕的研究成果，验证了基于深度学习的特定辐射源识别方法的有效性。数据是深度学习的三大支柱之一，这意味着基于深度学习的特定辐射源识别方法在训练阶段通常也需要海量、高质量电磁信号样本，但是在复杂多变的电磁环境中获取海量、高质量的电磁信号样本，并对其进行标注是十分困难并且成本高昂的。为此，部分研究者将目光聚焦在小样本特定辐射源识别问题上。为了揭示面向特定辐射源识别的小样本学习方法现有水平，推动小样本场景下的特定辐射源识别方法的研究与发展，文章系统地调研并综述了近年来国内外面向特定辐射源识别的小样本学习方法。首先，将已有的小样本特定辐射源识别方法根据有无辅助电磁信号样本对其进行分类，并进行问题描述；然后，分别综述已有的研究工作；最后，进行相关实验仿真并分析了存在的问题与未来的方向。
Abstract：With the proliferation of technologies such as 5th-generation mobile communication and the Internet of things, there has been an explosive growth in the number and types of emitters. Consequently, the recognition performance of traditional specific emitter identification(SEI) technology, which relies on statistical features and machine learning, is challenging to achieve practical application levels. Inspired by the remarkable performance of deep learning in fields like computer vision and natural language processing, numerous researchers have applied deep learning to SEI, yielding fruitful research outcomes that substantiate the effectiveness of deep learning-based SEI methods. Data is one of the fundamental pillars of deep learning, implying that deep learning-based SEI methods typically require extensive and high-quality electromagnetic signal samples for the training process. However, obtaining and accurately labeling massive and high-quality electromagnetic signal samples in complex and dynamic electromagnetic environments is difficult and expensive. Consequently, some researchers have focused on the problem of few-shot learning for specific emitter identification. To elucidate the current status of few-shot learning-based SEI methods and facilitate research and development in the context of few-shot scenarios, this paper comprehensively investigates and summarizes recent few-shot learning-based SEI methods conducted both domestically and internationally. Firstly, the existing few-shot learning-based SEI methods are classified based on whether auxiliary electromagnetic signal samples are utilized in the training process, accompanied by a description of the corresponding problem formulations. Subsequently, a summary of the existing research work is presented, followed by relevant experimental simulations and an analysis of the existing challenges and future directions.
[1] XU Q, ZHENG R, SAAD W, et al. Device fingerprinting in wireless networks:challenges and opportunities[J]. IEEE Communications Surveys and Tutorials, 2016, 18(1):94-104.
[2] XIE J D, LIU C, LIANG Y C, et al. Activity pattern aware spectrum sensing:a CNN-based deep learning approach[J]. IEEE Communications Letters, 2019, 23(6):1025-1028.
[3] GOK G, ALP Y K, ARIKAN O. A new method for specific emitter identification with results on real radar measurements[J]. IEEE Transactions on Information Forensics and Security, 2020, 15:3335-3346.
[4]杨睿.通信辐射源个体识别技术研究[D].成都：电子科技大学，2021.YANG R. Research on individual identification technology of communication radiator[D]. Chengdu:University of Electronic Science and Technology of China, 2021.(in Chinese)
[5]牛浩楠，王文灿，刘清波.基于卷积神经网络的雷达辐射源识别[J].现代防御技术，2021, 49(3):130-136.NIU H N, WANG W C, LIU Q B. Radar radiation source identification based on convolution neural network[J]. Modern Defence Technology, 2021, 49(3):130-136.(in Chinese)
[6] PATIL P W, BIRADAR K M, DUDHANE A, et al. An end-to-end edge aggregation network for moving object segmentation[C]//Proceedings of the 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition(CVPR), June 13-19, 2020, Seattle, WA, USA. New York:IEEE Xplore, 2020:8146-8155.
[7] LEE J, KIM E, YOON S. Anti-adversarially manipulated attributions for weakly and semi-supervised semantic segmentation[C]//Proceedings of the 2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition(CVPR),June 20-25, 2021, Nashville, TN, USA. New York:IEEE Xplore, 2021:4070-4078.
[8] TAIGMAN Y, YANG M, RANZATO M, et al. DeepFace:closing the gap to human-level performance in face verification[C]//Proceedings of the 2014 IEEE Conference on Computer Vision and Pattern Recognition(CVPR), June23-28, 2014, Columbus, OH, USA. New York:IEEE Xplore, 2014:1701-1708.
[9] YANG H, YUAN C F, ZHANG L, et al. STA-CNN:convolutional spatial-temporal attention learning for action recognition[J]. IEEE Transactions on Image Processing,2020, 29:5783-5793.
[10] JAJOO G, KUMAR Y, YADAV S K. Blind signal PSK/QAM recognition using clustering analysis of constellation signature in flat fading channel[J]. IEEE Communications Letters, 2019, 23(10):1853-1856.
[11] CORDOS C, MIHAIL?L, FARAGóP, et al. ECG signal classification using convolutional neural networks for biometric identification[C]//Proceedings of the 2021 44th International Conference on Telecommunications and Signal Processing(TSP), July 26-28, 2021, Brno, Czech Republic. New York:IEEE Xplore, 2021:167-170.
[12] XIA Y, MA Z Y, HUANG Z. Over-the-air radar emitter signal classification based on SDR[C]//Proceedings of the2021 6th International Conference on Intelligent Computing and Signal Processing(ICSP), April 9-11, 2021, Xi′an,China. New York:IEEE Xplore, 2021:403-408.
[13] ZHOU H J, JIAO L C, ZHENG S L, et al. Generative adversarial network-based electromagnetic signal classification:a semi-supervised learning framework[J]. China Communications, 2020, 17(10):157-169.
[14] MERCHANT K, REVAY S, STANTCHEV G, et al. Deep learning for RF device fingerprinting in cognitive communication networks[J]. IEEE Journal of Selected Topics in Signal Processing, 2018, 12(1):160-167.
[15] YU J B, HU A Q, LI G Y, et al. A robust RF fingerprinting approach using multisampling convolutional neural network[J]. IEEE Internet of Things Journal, 2019, 6(4):6786-6799.
[16] YANG Z Y, FU X, GUI G, et al. Rogue emitter detection using hybrid network of denoising autoencoder and deep metric learning[EB/OL].[2022-08-24]. https：//arxiv.org/abs/2212.00242.
[17] FU X, PENG Y, LIU Y C, et al. Semi-supervised specific emitter identification method using metric-adversarial training[J]. IEEE Internet of Things Journal, 2023, 10(12):10778-10789.
[18] 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.
[19] 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, 2020, 69(1):1091-1095.
[20] PENG Y, LIU P F, WANG Y, et al. Radio frequency fingerprint identification based on slice integration cooperation and heat constellation trace figure[J]. IEEE Wireless Communications Letters, 2022, 11(3):543-547.
[21] MERCHANT K, NOUSAIN B. Enhanced RF fingerprinting for IoT devices with recurrent neural networks[C]//Proceedings of the 2019 IEEE Military Communications Conference(MILCOM), November 12-14, 2019, Norfolk, VA,USA. New York:IEEE Xplore, 2020:590-597.
[22] WANG X L, ZHANG Y F, ZHANG H X, et al. Radio frequency signal identification using transfer learning based on LSTM[J]. Circuits, Systems, and Signal Processing,2020, 39(11):5514-5528.
[23] 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.
[24] YANG Y X, YAN T F. Radio frequency fingerprint recognition method based on generative adversarial net[C]//Proceedings of the 2021 13th International Conference on Communication Software and Networks(ICCSN), June 4-7,2021, Chongqing, China. New York:IEEE Xplore, 2021:361-364.
[25]陈浩，杨俊安，刘辉.基于深度残差适配网络的通信辐射源个体识别[J].系统工程与电子技术，2021, 43(3):603-609.CHEN H, YANG J A, LIU H. Communication transmitter individual identification based on deep residual adaptation network[J]. Systems Engineering and Electronics, 2021,43(3):603-609.(in Chinese)
[26]普运伟，刘涛涛，吴海潇，等.基于卷积双向长短时记忆网络的雷达辐射源信号识别[J].激光与光电子学进展，2022, 59(22):361-368.PU Y W, LIU T T, WU H X, et al. Radar emitter signal recognition based on convolutional bidirectional long-and short-term memory network[J]. Laser and Op toelectronics Progress, 2022, 59(22):361-368.(in Chinese)
[27]苏丹宁，曹桂涛，王燕楠，等.小样本雷达辐射源识别的深度学习方法综述[J].计算机科学，2022, 49(7):226-235.SU D N, CAO G T, WANG Y N, et al. Survey of deep learning for radar emitter identification based on small sample[J]. Computer Science, 2022, 49(7):226-235.(in Chinese)
[28] LIU C, FU X, GE Y L, et al. A robust few-shot SEI method using class-reconstruction and adversarial training[C]//Proceedings of the 2022 IEEE 96th Vehicular Technology Conference(VTC2022-Fall), September 26-29,2022, London, United Kingdom. New York:IEEE Xplore,2023:1-5.
[29] WANG C, FU X, WANG Y, et al. Few-shot specific emitter identification via hybrid data augmentation and deep metric learning[EB/OL].[2022-06-24]. https：//arxiv.org/abs/2212. 00252.
[30] HUANG L, PAN W J, ZHANG Y, et al. Data augmentation for deep learning-based radio modulation classification[J]. IEEE Access, 2019, 8:1498-1506.
[31] CAI Z R, MA W X, WANG X R, et al. The performance analysis of time series data augmentation technology for small sample communication device recognition[J]. IEEE Transactions on Reliability, 2023, 72(2):574-585.
[32] ZHANG W, JIANG L, DING C Z, et al. Small sample identification for specific emitter based on adversarial embedded networks[C]//Proceedings of the 11th International Conference on Image and Graphics, ICIG 2021, August6-8, 2021, Haikou, China. Cham:Springer, 2021:226-237.
[33] XU X D, JIANG T, GONG J L, et al. WLAN interference signal recognition using an improved quadruple generative adversarial network[J]. Digital Signal Processing, 2021,117:103188.
[34]丁辰伟，孙闽红，官友廉.基于小样本条件的雷达辐射源个体识别算法[J].弹箭与制导学报，2021, 41(3):25-30.DING C W, SUN M H, GUAN Y L. Algorithm of specific radar emitter identification based on few-shot learning[J].Journal of Projectiles, Rockets, Missiles and Guidance,2021, 41(3):25-30.(in Chinese)
[35]梁先明.一种优化孪生网络的小样本辐射源个体识别方法[J].电讯技术，2022, 62(6):695-701.LIANG X M. An emitter individual identification method for small samples based on optimized siamese networks[J].Telecommunication Engineering, 2022, 62(6):695-701.(in Chinese)
[36] QIAN Y H, QI J, KUAI X Y, et al. Specific emitter identification based on multi-level sparse representation in automatic identification system[J]. IEEE Transactions on Information Forensics and Security, 2021, 16:2872-2884.
[37] HUANG K J, YANG J A, LIU H, et al. Deep learning of radio frequency fingerprints from limited samples by masked autoencoding[J/OL]. IEEE Wireless Communications Letters, 2022[2022-08-24]. http：//ieee xplore.ieee.org/stamp/stamp.jsp? tp=&arnumber=9802743. DOI:10. 1109/IWC. 2022. 3184674.
[38] XING E P, NG A Y, JORDAN M I, et al. Distance metric learning, with application to clustering with side-information[C]//Proceedings of the 15th International Conference on Neural Information Processing Systems, New York:ACM,2002:521-528.
[39] BERTINETTO L, VALMADRE J, HEN RIQUES J F, et al. Fully-convolutional Siamese networks for object tracking[M]//Lecture Notesin Computer Science. Cham:Springer International Publishing, 2016:850-865.
[40] HOFFER E, AILON N. Deep metric learning using triplet network[M]//Similarity-Based Pattern Recognition. Cham:Springer International Publishing, 2015:84-92.
[41] WEN Y D, ZHANG K P, LI Z F, et al. A discriminative feature learning approach for deep face recognition[C]//Proceedings of the 14th European Conference on Computer Vision, October 11-14, 2016, Amsterdam, The Netherlands. Cham:Springer, 2016:499-515.
[42] WANG Y, GUI G, LIN Y, et al. Few-shot specific emitter identification via deep metric ensemble learning[J]. IEEE Internet of Things Journal, 2022, 9(24):24980-24994.
[43] MELGOZA C M, GROOM T, LIN H, et al. Environment classification and deinterleaving using Siamese networks and few-shot learning[C]//Proceedings of the 2021 IEEE12th Annual Ubiquitous Computing, Electronics and Mobile Communication Conference(UEMCON), December 1-4, 2021, New York, NY, USA. New York:IEEE Xplore,2022:555-559.
[44] HUANG J, WU B, LI P, et al. Few-shot learning for radar emitter signal recognition based on improved prototypical network[J]. Remote Sensing, 2022, 14(7):1681.
[45] ZHANG X X, WANG Y, ZHANG Y B, et al. Data augmentation aided few-shot learning for specific emitter identification[C]//Proceedings of the 2022 IEEE 96th Vehicular Technology Conference(VTC 2022-Fall), September26-29, 2022, London, United Kingdom. New York:IEEE Xplore, 2023:1-5.
[46] YANG N, ZHANG B N, DING G R, et al. Specific emitter identification with limited samples:a model-agnostic meta-learning approach[J]. IEEE Communications Letters,2022, 26(2):345-349.
[47] XIE C X, ZHANG L M, ZHONG Z G. Few-shot specific emitter identification based on variational mode decomposition and meta-learning[J]. Wireless Communications and Mobile Computing, 2022, 2022:1-9.
[48] XIE C X, ZHANG L M, ZHONG Z G. Few-shot unsupervised specific emitter identification based on density peak clustering algorithm and meta-learning[J]. IEEE Sensors Journal, 2022, 22(18):18008-18020.
[49] XIE C X, ZHANG L M, ZHONG Z G. Meta learningbased open-set identification system for specific emitter identification in non-cooperative scenarios[J]. KSII Transactions on Internet and Information Systems, 2022, 16(5):1755-1777.
[50] TU Y, LIN Y, ZHA H R, et al. Large-scale real-world radio signal recognition with deep learning[J]. Chinese Journal of Aeronautics, 2022, 35(9):35-48.
基本信息:
DOI：10.12194/j.ntu.20220928001
中图分类号:TN957.51
引用信息:
[1]桂冠,陶梦圆,王诚等.面向特定辐射源识别的小样本学习方法综述[J],2023,22(03):1-16.DOI:10.12194/j.ntu.20220928001.
基金信息:
科技创新2030——“新一代人工智能”重大项目（2021ZD0113003）