基于生成对抗网络的雷达脉内信号去噪与识别

doi:10.19665/j.issn1001-2400.20230312

摘要/Abstract

摘要：

近年来,深度神经网络在计算机视觉等领域取得了突破性进展,然而在射频信号处理领域,如电子支援侦察系统中的雷达辐射源识别任务,相关技术的发展仍处于起步阶段。在实际军事应用场景中,噪声的存在是影响深度神经网络性能发挥的关键因素。例如,在高信噪比环境下训练至收敛的深度模型分类器在处理低信噪比数据时往往性能下降严重。为了解决上述问题,提出了一种生成对抗式的去噪网络,实现了端到端的雷达信号去噪和脉内调制类型识别。该模型由生成器、鉴别器和分类器三部分组成,其中,生成器为编解码器结构,通过对称的上采样和下采样操作提取输入雷达信号中高阶特征向量,从噪声中恢复出干净信号;鉴别器则用来判断生成器输出去噪结果的真伪;在此基础上,将分类器与上述两者级联,使得去噪结果符合分类所需的语义信息。实验结果表明,所提算法在密集噪声环境下具备高质量的信号去噪效果和较高的分类准确度;与已有算法相比,算法在低信噪比环境数据上的迁移能力具有一定的优越性。

关键词: 雷达辐射源, 信号识别, 卷积神经网络, 生成对抗网络, 信号去噪

Abstract:

While deep neural networks have achieved an impressive success in computer vision,the related research remains embryonic in radio frequency signal processing,i.e.,a vital task in modern wireless systems,for example,the electronic reconnaissance system.Noise corruption is a harmful but unavoidable factor causing severe performance degradation in the signal processing procedure,and thus has persistently been an intractable problem in the radio frequency domain.For example,a classifier trained on the high signal-to-noise ratio(SNR) data might experience a severe performance degradation when dealing with low SNR data.To address this problem,in this paper we leverage the powerful data representation capacity of deep learning and propose a Generative Adversarial Denoising and classification Network(GADNet) for radar signal restoration and a classification task.The proposed GADNet consists of a generator,a discriminator and a classifier fulfilling an end-to-end workflow.The encoder-decoder structure generator is trained to extract the high-level features and recover signals.Meanwhile,it fools the discriminator’s judges by bewildering the denoising results coming from the clean data.The classification loss from the classifier is adopted jointly to the training procedure.Extensive experiments demonstrate the benefit of the proposed technique in terms of high-quality restoration and accurate classification for radar signals with intense noise.Moreover,it also exhibits superior transferability in low SNR environments compared to the state-of-the-art methods.

Key words: radar emitter, signal recognition, convolutional neural networks, generative adversarial network, signal denoising

中图分类号:

TN971

杜明洋, 杜蒙, 潘继飞, 毕大平. 基于生成对抗网络的雷达脉内信号去噪与识别[J]. 西安电子科技大学学报, 2023, 50(6): 133-147.

DU Mingyang, DU Meng, PAN Jifei, BI Daping. Generative adversarial model for radar intra-pulse signal denoising and recognition[J]. Journal of Xidian University, 2023, 50(6): 133-147.

图/表 16

图1

图2

图3

图4

表1

表2

图5

图6

图7

图8

表3

图9

表4

图10

表5

图11

参考文献 38

[1]	RIYAZ S, SANKHE K, IOANNIDIS S, et al. Deep Learning Convolutional Neural Networks for Radio Identification[J]. IEEE Communications Magazine, 2018, 56(9):146-152.
[2]	LIU Z M. Recognition of Multi-Function Radars via Hierarchically Mining and Exploiting Pulse Group Patterns[J]. IEEE Transactions on Aerospace and Electronic Systems, 2020, 44(6):4659-4672.
[3]	孙丽婷, 黄知涛, 王翔, 等. 辐射源指纹特征提取方法述评[J]. 雷达学报, 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.
[4]	刘明骞, 孟燕, 张卫东. 雷达辐射源识别的效能综合评估方法[J]. 西安电子科技大学学报, 2019, 46(6):1-8.
	LIU Mingqian, Meng Yan, ZHANG Weidong. Method for Comprehensive Evalution of Effectiveness of Radar Emitter Signals Recognition[J]. Journal of Xidian University, 2019, 46(6):1-8.
[5]	ANJANEYULU L, MURTHY N S, SARMA N. Radar Emitter Classification Using Self-Organising Neural Network Models[C]// 2008 International Conference on Recent Advances in Microwave Theory and Applications.Piscataway:IEEE, 2008:431-433.
[6]	REDMON J, DIVVALA S, GIRSHICK R, et al. You Only Look Once:Unified,Real-Time Object Detection[C]// Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition.Piscataway:IEEE, 2016:779-788.
[7]	FUAD M T H, FIME A A, SIKDER D, et al. Recent Advances in Deep Learning Techniques for Face Recognition[J]. IEEE Access, 2021, 9:99112-99142. doi: 10.1109/ACCESS.2021.3096136
[8]	ZHANG Z, GEIGER J, POHJALAINEN J, et al. Deep Learning for Environmentally Robust Speech Recognition:An Overview of Recent Developments[J]. ACM Transactions on Intelligent Systems and Technology(TIST), 2018, 9(5):1-28.
[9]	ZHANG H, ZHOU F, WU Q, et al. A Novel Automatic Modulation Classification Scheme Based on Multi-Scale Networks[J]. IEEE Transactions on Cognitive Communications and Networking, 2021, 8(1):97-110. doi: 10.1109/TCCN.2021.3091730
[10]	HUANG S, JIANG Y, GAO Y, et al. Automatic Modulation Classification Using Contrastive Fully Convolutional Network[J]. IEEE Wireless Communications Letters, 2019, 8(4):1044-1047. doi: 10.1109/LWC.5962382
[11]	DU M, HE X, CAI X, et al. Balanced Neural Architecture Search and its Application in Specific Emitter Identification[J]. IEEE Transactions on Signal Processing, 2021, 69:5051-5065. doi: 10.1109/TSP.2021.3107633
[12]	O’SHEA T J, CORGAN J, CLANCY T C. Convolutional Radio Modulation Recognition Networks[C]// International Conference on Engineering Applications of neural networks.Heidelberg:Springer, 2016:213-226.
[13]	O’SHEA T J, WEST N. Radio Machine Learning Dataset Generation with GNU Radio[C]// Proceedings of the GNU Radio Conference.Boulder:GNU, 2016:1-6.
[14]	HAUSER S C, HEADLEY W C, MICHAELS A J. Signal Detection Effects on Deep Neural Networks Utilizing Raw IQ for Modulation Classification[C]// MILCOM 2017-2017 IEEE Military Communications Conference(MILCOM).Piscataway:IEEE, 2017:121-127.
[15]	YILDIRIM A, KIRANYAZ S. 1D Convolutional Neural Networks Versus Automatic Classifiers for Known LPI Radar Signals under White Gaussian Noise[J]. IEEE Access, 2020, 8:180534-180543. doi: 10.1109/Access.6287639
[16]	O’SHEA T J, ROY T, CLANCY T C. Over-the-Air Deep Learning Based Radio Signal Classification[J]. IEEE Journal of Selected Topics in Signal Processing, 2018, 12(1):168-179. doi: 10.1109/JSTSP.2018.2797022
[17]	HUYNH T T, HUA C H, PHAM Q V, et al. MCNet:An Efficient CNN Architecture for Robust Automatic Modulation Classification[J]. IEEE Communications Letters, 2020, 24(4):811-815. doi: 10.1109/COML.4234
[18]	SZEGEDY C, LIU W, JIA Y, et al. Going Deeper with Convolutions[C]// Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition.Piscataway:IEEE, 2015:1-9.
[19]	ZHANG H, YUAN L, WU G, et al. Automatic Modulation Classification Using Involution Enabled Residual Networks[J]. IEEE Wireless Communications Letters, 2021, 10(11):2417-2420. doi: 10.1109/LWC.2021.3102069
[20]	LI D, HU J, WANG C, et al. Involution:Inverting the Inherence of Convolution for Visual Recognition[C]// Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition.Piscataway:IEEE, 2021:12321-12330.
[21]	HE K, ZHANG X, REN S, et al. Deep Residual Learning for Image Recognition[C]// Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition.Piscataway:IEEE, 2016:770-778.
[22]	WEST N E, O'SHEA T. Deep Architectures for Modulation Recognition[C]// 2017 IEEE International Symposium on Dynamic Spectrum Access Networks(DySPAN).Piscataway:IEEE, 2017:1-6.
[23]	DU M, ZHONG P, CAI X, et al. DNCNet:Deep Radar Signal Denoising and Recognition[J]. IEEE Transactions on Aerospace and Electronic Systems, 2022, 58(4):3549-3562. doi: 10.1109/TAES.2022.3153756
[24]	HU S, PEI Y, LIANG P P, et al. Deep Neural Network for Robust Modulation Classification under Uncertain Noise Conditions[J]. IEEE Transactions on Vehicular Technology, 2019, 69(1):564-577. doi: 10.1109/TVT.25
[25]	刘松涛, 赵帅, 汪慧阳. 雷达辐射源识别技术新进展[J]. 中国电子科学研究院学报, 2022, 17(6):523-533.
	LIU Songtao, ZHAO Shuai, WANG Huiyang. New Development on the Technology of Radar Emitter Identification[J]. Journal of CAEIT, 2022, 17(6):523-533.
[26]	GOODFELLOW I, POUGET-ABADIE J, MIRZA M, et al. Generative Adversarial Networks[J]. Communications of the ACM, 2020, 63(11):139-144. doi: 10.1145/3422622
[27]	LIU D, WEN B, JIAO J, et al. Connecting Image Denoising and High-Level Vision Tasks via Deep Learning[J]. IEEE Transactions on Image Processing, 2020, 29:3695-3706. doi: 10.1109/TIP.83
[28]	赵集. Alpha稳定分布环境下自适应滤波算法研究[D]. 成都: 电子科技大学, 2020.
[29]	于浩洋, 尹良, 李书芳. 生成对抗网络小样本雷达调制信号识别算法[J]. 西安电子科技大学学报, 2021, 48(6):96-104.
	YU Haoyang, YIN Liang, LI Shufang, et al. Recognition Algorithm for the Little Sample Radar Modulation Signal Based on the Generative Adversarial Network[J]. Journal of Xidian University, 2021, 48(6):96-104.
[30]	LECUN Y, BENGIO Y, HINTON G. Deep Learning[J]. Nature, 2015, 521(7553):436-444. doi: 10.1038/nature14539
[31]	SUN J, XU G, REN W, et al. Radar Emitter Classification Based on Unidimensional Convolutional Neural Network[J]. IET Radar,Sonar & Navigation, 2018, 12(8):862-867. doi: 10.1049/rsn2.v12.8
[32]	BROOKS T, MILDENHALL B, XUE T, et al. Unprocessing Images for Learned Raw Denoising[C]// Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition.Piscataway:IEEE, 2019:11036-11045.
[33]	ARJOVSKY M, CHINTALA S, BOTTOU L. Wasserstein Generative Adversarial Networks[C]// International Conference on Machine Learning(PMLR). New York: ACM, 2017:214-223.
[34]	LUKE M, BEN P, DAVID P, et al. Unrolled Generative Adversarial Networks(2017)[J/OL].[2017-05-12]. https://arxiv.org/abs/1611.02163.
[35]	GULRAJANI I, AHMED F, ARJOVSKY M, et al. Improved Training of Wasserstein GANs[C]// Proceedings of the 31st International Conference on Neural Information Processing Systems. New York: ACM, 2017:5769-5779.
[36]	SELVARAJU R R, COGSWELL M, DAS A, et al. Grad-CAM:Visual Explanations from Deep Networks via Gradient-Based Localization[C]// Proceedings of the IEEE International Conference on Computer Vision.Piscataway:IEEE, 2017:618-626.
[37]	RONNEBERGER O, FISCHER P, BROX T. U-Net:Convolutional Networks for Biomedical Image Segmentation[C]// International Conference on Medical Image Computing and Computer-Assisted Intervention.Heidelberg:Springer, 2015:234-241.
[38]	DIAKOGIANNIS F I, WALDNER F, CACCETTA P, et al. ResUNet-a:A Deep Learning Framework for Semantic Segmentation of Remotely Sensed Data[J]. ISPRS Journal of Photogrammetry and Remote Sensing, 2020, 162:94-114. doi: 10.1016/j.isprsjprs.2020.01.013

LFM/SFM	脉冲宽度/μs	带宽/MHz
1	1~16	1
2	1~16	5
3	10	20
4	1~20	180

Frank/P1/P2/ P3/P4	码元持续时间 /μs	编码位数
1	1	64
2	1.5	64
3	2	64

模型	训练阶段Ⅰ	训练阶段Ⅱ	测试准确度		平均值
模型	训练阶段Ⅰ	训练阶段Ⅱ	-5 dB	-10 dB	平均值
UNet	99.80	99.82	85.45	51.76	79.01
ResUNet	99.89	99.90	92.87	69.86	87.54
GC	99.95	99.98	93.20	90.10	94.43
GN	99.97	99.99	95.00	52.30	82.43

模型	噪声类型	训练阶段Ⅰ	训练阶段Ⅱ	测试准确度		平均值
模型	噪声类型	训练阶段Ⅰ	训练阶段Ⅱ	1 dB	-1 dB	平均值
	脉冲噪声	43.27	77.25	24.24	5.92	35.80
UNet	高斯白噪声	71.50	83.83	56.49	40.84	60.40
	高斯色噪声	67.20	83.85	51.55	33.96	56.50
	脉冲噪声	47.80	79.78	20.92	9.44	36.70
ResUNet	高斯白噪声	77.27	81.89	56.22	43.39	60.50
	高斯色噪声	80.70	82.14	58.74	45.27	62.20
	脉冲噪声	76.60	82.50	23.00	9.50	38.30
GC	高斯白噪声	71.00	85.70	63.50	52.30	67.20
	高斯色噪声	71.40	85.70	61.70	49.70	65.70
	脉冲噪声	54.30	74.50	20.10	7.30	34.00
GN	高斯白噪声	83.30	85.40	56.40	41.30	61.00
	高斯色噪声	82.90	85.20	56.90	41.40	61.20

模型	5 dB	1 dB	-1 dB	平均值
UNet	无法收敛
ResUNet	76.21	71.91	25.50	57.90
GC	81.10	77.20	26.20	61.50
GN	75.10	71.00	19.80	55.30