对调制识别网络隐身的雷达发射信号生成方法

doi:10.12305/j.issn.1001-506X.2024.07.09

摘要/Abstract

摘要：

雷达对抗场景中, 电子侦察系统通过引入基于深度学习方法的智能脉冲调制识别网络, 极大提升了对雷达信号的识别准确率。为了提高雷达信号的调制隐身抗识别能力, 提出一种可以令深度识别网络错误预测的雷达发射信号生成方法。该方法首先通过短时傅里叶变换得到信号的时频谱; 然后迭代生成携带调制隐身信息的时频谱; 最后利用改进逆短时傅里叶变换得到时域调制隐身发射信号。该方法生成的雷达信号对以时频图为输入的调制识别网络隐身, 并可实现回波信号的脉冲压缩处理。仿真结果验证了所生成信号的抗识别有效性、噪声鲁棒性和脉压可行性。

关键词: 雷达发射信号, 时频分析, 自动调制分类, 射频隐身, 深度神经网络

Abstract:

The electronic reconnaissance system in radar countermeasure scenarios greatly improves the recognition accuracy of radar signals by introducing an intelligent pulse modulation recognition network based on deep learning methods. In order to improve the modulation stealth and anti-recognition ability of radar signals, a radar transmission signal generation method that can make the deep recognition network make incorrect predictions is proposed. Firstly, the time-frequency spectrum of the signal is obtained through short-time Fourier transform (STFT). Then, a time-frequency spectrum carrying modulated stealth information is generated iteratively. Finally, the improved inverse STFT is used to obtain the time-domain modulated stealth transmission signal. The radar signal generated by the proposed method is invisible to the modulation recognition network input from the time-frequency map, and can achieve pulse compression processing of the echo signal. The simulation results verified the effectiveness of the generated signal in resisting recognition, noise robustness, and pulse compression feasibility.

Key words: radar transmitting signal, time-frequency analysis, automatic modulation classification, radio frequency stealth, deep neural network (DNN)

中图分类号:

TN957

张瑞斌, 朱梦韬, 李云杰. 对调制识别网络隐身的雷达发射信号生成方法[J]. 系统工程与电子技术, 2024, 46(7): 2256-2268.

Ruibin ZHANG, Mengtao ZHU, Yunjie LI. Radar transmitting signal generation method for modulation recognition network stealth[J]. Systems Engineering and Electronics, 2024, 46(7): 2256-2268.

图/表 17

图1

表1

图2

图3

图4

图5

图6

表2

图7

表3

图8

图9

表4

图10

表5

表6

图11

参考文献 39

1	王谦喆, 何召阳, 宋博文, 等. 射频隐身技术研究综述[J]. 电子与信息学报, 2018, 40 (6): 1505- 1514.
	WANG Q Z , HE Z Y , SONG B W , et al. Overview on RF stealth technology research[J]. Journal of Electronics & Information Technology, 2018, 40 (6): 1505- 1514.
2	樊依晨. 机载战场侦察雷达射频隐身波形设计[D]. 北京: 中国电子科技集团公司电子科学研究院, 2021.
	FAN Y C. Research on RF stealth waveform design of airborne battlefield surveillance radar[D]. Beijing: China Academic of Electronics and Information Technology, 2021.
3	WANG W Q . Moving-target tracking by cognitive RF stealth radar using frequency diverse array antenna[J]. IEEE Trans. on Geoscience and Remote Sensing, 2016, 54 (7): 3764- 3773. doi: 10.1109/TGRS.2016.2527057
4	YANG H B, ZHOU J J, WANG F, et al. Design and analysis of Costas/PSK RF stealth signal waveform[C]//Proc. of the IEEE/CIE International Conference on Radar, 2011: 1247-1250.
5	杨宇晓, 左瑞芹. 基于混沌序列的射频隐身跳频周期设计方法[J]. 航空兵器, 2016, (5): 34- 38.
	YANG Y X , ZUO R Q . A design method for hopping cycle of RF stealth based on chaotic sequences[J]. Aero Weaponry, 2016, (5): 34- 38.
6	王志涛. 基于MIMO-OFDM的低截获概率雷达波形设计方法与仿真实现[D]. 西安: 西安电子科技大学, 2017.
	WANG Z T. MIMO-OFDM waveform design and simulation of LPI radar[D]. Xi'an: Xidian University, 2017.
7	YANG Y X , WANG D X , HUANG Q . Design method of radio frequency stealth frequency hopping communications based on four-dimensional hyperchaotic system[J]. Journal of Astronautics, 2020, 41 (10): 1341- 1349.
8	JIA J W , HAN Z Z , LIANG Y Y , et al. Design of multi-parameter compound modulated RF stealth anti-sorting signals based on hyperchaotic interleaving feedback[J]. Entropy, 2022, 24 (9): 1283- 1314. doi: 10.3390/e24091283
9	孙岩博, 王亚涛, 黄小艳. 基于随机化调制的射频隐身波形识别性能分析[J]. 计算机仿真, 2020, 37 (10): 1-5, 13.
	SUN Y B , WANG Y T , HUANG X Y . Recognition performance analysis of RF stealth waveform based on randomized modulations[J]. Computer Simulation, 2020, 37 (10): 1-5, 13.
10	付银娟, 李勇, 黄琼丹, 等. OFD-NLFM雷达信号的优化设计及射频隐身性能分析[J]. 系统工程与电子技术, 2017, 39 (10): 2209- 2214. doi: 10.3969/j.issn.1001-506X.2017.10.08
	FU Y J , LI Y , HUANG Q D , et al. Optimized design of OFD-NLFM radar signals and analysis of RF stealth performance[J]. Systems Engineering and Electronics, 2017, 39 (10): 2209- 2214. doi: 10.3969/j.issn.1001-506X.2017.10.08
11	路晴辉. 机载认知雷达射频隐身波形设计及处理方法[D]. 成都: 电子科技大学, 2022.
	LU Q H. Radio frequency stealth waveform design and processing method for airborne cognitive radar[D]. Chengdu: University of Electronic Science and Technology of China, 2022.
12	LI J C, QI L, LIN Y. Research on modulation identification of digital signals based on deep learning[C]//Proc. of the IEEE International Conference on Electronic Information and Communication Technology, 2016: 402-405.
13	WANG X B, HUANG G M, ZHOU Z W, et al. Radar emitter recognition based on the short time Fourier transform and convolutional neural networks[C]//Proc. of the 10th International Congress on Image and Signal Processing, BioMedical Engineering and Informatics, 2017: 1-5.
14	ZHANG M , DIAO M , GUO L M . Convolutional neural networks for automatic cognitive radio waveform recognition[J]. IEEE Access, 2017, 5, 11074- 11082. doi: 10.1109/ACCESS.2017.2716191
15	QU Z Y , MAO X J , DENG Z A . Radar signal intra-pulse modulation recognition based on convolutional neural network[J]. IEEE Access, 2018, 6, 43874- 43884. doi: 10.1109/ACCESS.2018.2864347
16	KONG S H , KIM M , HOANG L M , et al. Automatic LPI radar waveform recognition using CNN[J]. IEEE Access, 2018, 6, 4207- 4219. doi: 10.1109/ACCESS.2017.2788942
17	SUN D C, CHEN Y, LIU J A, et al. Digital signal modulation recognition algorithm based on VGGNet model[C]//Proc. of the IEEE 5th International Conference on Computer and Communication, 2019: 1575-1579.
18	YU Z Y , TANG J L , WANG Z . GCPS: a CNN performance evaluation criterion for radar signal intrapulse modulation recognition[J]. IEEE Communications Letters, 2021, 25 (7): 2290- 2294. doi: 10.1109/LCOMM.2021.3070151
19	PAN Z S , WANG S F , ZHU M T , et al. Automatic waveform recognition of overlapping LPI radar signals based on multi-instance multi-label learning[J]. IEEE Signal Processing Letters, 2020, 27 (7): 1275- 1279.
20	JIANG W K , LI Y , LIAO M M , et al. An improved LPI radar waveform recognition framework with LDC-Unet and SSR-Loss[J]. IEEE Signal Processing Letters, 2021, 29 (11): 149- 153.
21	PAN Z S , WANG S F , LI Y J . Residual attention aided U-net GAN and multi-instance multi-label classifier for automatic waveform recognition of overlapping LPI radar signals[J]. IEEE Trans. on Aerospace and Electronic Systems, 2022, 58 (5): 4377- 4395. doi: 10.1109/TAES.2022.3160978
22	HE K M, ZHANG X Y, REN S Q, et al. Deep residual learning for image recognition[C]//Proc. of the IEEE Conference on Computer Vision and Pattern Recognition, 2016: 770-778.
23	SIMONYAN K, ZISSERMAN A. Very deep convolutional networks for large-scale image recognition[EB/OL]. [2023-03-02]. https://arxiv.org/abs/1409.1556.
24	SZEGEDY C, ZAREMBA W, SUTSKEVER I, et al. Intriguing properties of neural networks[EB/OL]. [2023-03-23]. https://arxiv.org/abs/1312.6199v1.
25	GOODFELLOW I J, SHLENS J, SZEGEDY C. Explaining and harnessing adversarial examples[EB/OL]. [2023-03-23]. https://arxiv.org/abs/1412.6572v1.
26	DU C , CONG Y L , ZHANG L , et al. A practical deceptive jamming method based on vulnerable location awareness adversarial attack for radar HRRP target recognition[J]. IEEE Trans. on Information Forensics and Security, 2022, 17 (5): 2410- 2424.
27	TANG S T, TAO M L, SU J, et al. Radio signal smart deception based on adversarial learning[C]//Proc. of the CIE International Conference on Radar, 2021: 2259-2262.
28	YUAN X J, CHEN Y X, ZHAO Y, et al. CommanderSong: a systematic approach for practical adversarial voice recognition[C]//Proc. of the 27th USENIX Security Symposium, 2018: 49-64.
29	SADEGHI M , LARSSON E G . Adversarial attacks on deep-learning based radio signal classification[J]. IEEE Wireless Communications Letters, 2018, 8 (1): 213- 216.
30	XIE Y, SHI C, LI Z H, et al. Real-time, universal, and robust adversarial attacks against speaker recognition systems[C]//Proc. of the IEEE International Conference on Acoustics, Speech and Signal Processing, 2020: 1738-1742.
31	ZHAO H J , TIAN Q , PAN L , et al. The technology of adversarial attacks in signal recognition[J]. Physical Communication, 2020, 43, 101199. doi: 10.1016/j.phycom.2020.101199
32	MCDANIEL P , PAPERNOT N , CELIK Z B . Machine learning in adversarial settings[J]. IEEE Security & Privacy, 2016, 14 (3): 68- 72.
33	MOOSAVI-DEZFOOLI S M, FAWZI A, FROSSARD P. Deepfool: a simple and accurate method to fool deep neural networks[C]//Proc. of the IEEE Conference on Computer Vision and Pattern Recognition, 2016: 2574-2582.
34	GABOR D . Theory of communication[J]. Journal of the Institution of Electrical Engineers-part Ⅲ: Radio and Communication Engineering, 1946, 93 (26): 429- 457.
35	LE-ROUX J, ONO N, SAGAYAMA S. Explicit consistency constraints for STFT spectrograms and their application to phase reconstruction[C]//Proc. of the ISCA Tutorial and Research Workshop on Statistical and Perceptual Audition, 2008: 23-28.
36	GRIFFIN D , LIM J . Signal estimation from modified short-time Fourier transform[J]. IEEE Trans. on Acoustics, Speech, and Signal Processing, 1984, 32 (2): 236- 243. doi: 10.1109/TASSP.1984.1164317
37	WANG Z , BOVIK A C , SHEIKH H R , et al. Image quality assessment: from error visibility to structural similarity[J]. IEEE Trans. on Image Processing,, 2004, 13 (4): 600- 612. doi: 10.1109/TIP.2003.819861
38	宣琦, 周晴, 崔慧, 等. 信号人工智能对抗攻击综合分析平台[J]. 信息安全学报, 2021, 6 (4): 141- 149.
	XUAN Q , ZHOU Q , CUI H , et al. A comprehensive evaluation platform of adversarial attacks on artificial intelligence for signal[J]. Journal of Cyber Security, 2021, 6 (4): 141- 149.
39	DE M A , HUANG Y , PIEZZO M , et al. Design of optimized radar codes with a peak to average power ratio constraint[J]. IEEE Trans. on Signal Processing, 2011, 59 (6): 2683- 2697. doi: 10.1109/TSP.2011.2128313

调制类型	f(n)/Hz	φ(n)/rad
Barker	f₀	0, π
Costas	f_i∈{f₁, f₂, …, f_M}	φ₀
LFM	f₀+nK/2	φ₀

雷达信号	参数	区间
Barker	载频f_c/MHz	U[f_s/8, f_s/4]
	码长N_c	{7, 11, 13}
	循环周期数cpp	U[1, 2]
	码周期N_p	U[200, 400]
Costas	载频f_c/MHz	U[f_s/20, f_s/10]
	频率切换数N	U[3, 6]
	基频f_min/MHz	f_s/25
LFM	初始频率f₀/MHz	U[f_s/16, f_s/8]
LFM	带宽B/MHz	U[f_s/8, f_s/6]

调制识别模型	VGG11	VGG13	VGG16	VGG19	ResNet18	ResNet34	ResNet50	CNN-base
准确率	98.75	99.31	98.75	99.17	94.58	98.06	96.39	97.64

模型	攻击成功率	-10 dB	-6 dB	-2 dB	0 dB	2 dB	6 dB	10 dB	理想信号
VGG16	ASR1	96.67	97.78	98.89	95.56	94.44	88.89	92.21	92.22
	ASR2	96.55	95.55	95.51	95.35	96.47	95.00	95.45	98.82
	ASR3	80.95	77.15	83.81	85.71	82.86	84.76	85.71	92.38
	ASR	75.55	72.08	79.16	78.10	75.49	71.58	75.44	84.19
ResNet18	ASR1	98.89	91.11	93.33	88.89	86.67	82.22	77.78	77.78
	ASR2	94.38	94.25	94.32	94.12	93.75	93.56	95.77	96.15
	ASR3	100.00	100.00	98.10	100.00	100.00	100.00	100.00	99.05
	ASR	93.33	85.87	86.36	83.66	81.25	76.93	74.49	74.08
CNN-base	ASR1	100.00	100.00	98.89	98.89	100.00	98.89	95.56	92.22
	ASR2	97.78	100.00	98.89	100.00	100.00	98.88	100.00	100.00
	ASR3	79.05	77.14	85.72	95.24	90.47	87.62	93.33	92.38
	ASR	77.30	77.14	83.83	94.18	90.47	85.68	89.19	85.19

模型	信号类型	-10 dB	-6 dB	-2 dB	0 dB	2 dB	6 dB	10 dB	理想信号
VGG16	Barker	6.05	5.72	5.32	5.02	4.46	3.54	2.86	1.67
	Costas	5.55	5.51	5.17	4.96	4.07	3.58	2.70	1.40
	LFM	5.78	5.82	5.12	5.04	4.69	3.92	2.93	1.40
ResNet18	Barker	6.21	5.87	5.43	5.02	3.25	3.47	2.76	2.37
	Costas	5.94	5.49	5.49	4.69	4.52	4.51	3.76	2.69
	LFM	5.88	5.79	5.35	5.66	5.08	4.35	4.73	3.77
CNN-base	Barker	6.08	5.60	4.89	5.16	4.46	3.54	3.11	5.28
	Costas	5.69	5.48	5.35	4.75	4.65	4.05	3.67	5.14
	LFM	5.83	5.49	5.04	5.52	4.74	3.78	3.76	7.14