基于含干扰数据的SAR图像域深度学习ISRJ抑制方法

doi:10.12305/j.issn.1001-506X.2026.10.13

系统工程与电子技术 ›› 2026, Vol. 48 ›› Issue (10): 3404-3416.doi: 10.12305/j.issn.1001-506X.2026.10.13

• 传感器与信号处理 • 上一篇

基于含干扰数据的SAR图像域深度学习ISRJ抑制方法

陈兴宏¹(), 张劲东¹(), 万程²()

1. 南京航空航天大学电子信息工程学院，江苏南京 211106
2. 南京航空航天大学人工智能学院，江苏南京 211106

收稿日期:2025-06-12 接受日期:2025-09-03 出版日期:2026-10-25 发布日期:2026-09-30
通讯作者: 万程 E-mail:18629423485@163.com;zjdjs@126.com;wanch@nuaa.edu.cn
作者简介:陈兴宏（2001—），男，硕士研究生，主要研究方向为智能图像处理
张劲东（1981—），男，教授，博士，主要研究方向为新体制雷达研发、雷达信号分析与处理、高速数字信号处理系统设计与实现
基金资助:
国家自然科学基金（62171220）资助课题

Deep learning method for ISRJ suppression in SAR image domain based on jammed data

Xinghong Chen¹(), Jindong Zhang¹(), Cheng Wan²()

1. College of Electronic and Information Engineering，Nanjing University of Aeronautics and Astronautics，Nanjing 211106，China
2. College of Artificial Intelligence，Nanjing University of Aeronautics and Astronautics，Nanjing 211106，China

Received:2025-06-12 Accepted:2025-09-03 Online:2026-10-25 Published:2026-09-30
Contact: Cheng Wan E-mail:18629423485@163.com;zjdjs@126.com;wanch@nuaa.edu.cn

摘要/Abstract

摘要：

针对间歇采样转发干扰（interrupted sampling repeater jamming，ISRJ）严重影响合成孔径雷达图像可靠性的问题，提出一种采用含ISRJ图像对的复数域深度学习合成孔径雷达ISRJ抑制方法，以及复数域深度残差收缩U型网络（complex-valued deep residual shrinkage U-Net，CV-DRSUNet）。通过构建复数域残差收缩模块，挖掘目标场景与干扰信号在合成孔径雷达复数域图像中的表征方式，实现ISRJ抑制。同时，提出一种不依赖无干扰场景图像的训练策略，一定程度上解决场景中存在干扰时干净图像难以获取的问题，扩展了深度学习在干扰抑制中的应用。通过数学推导和多组实验，验证了所提训练策略与网络模型在ISRJ抑制方面的有效性，证明所提方法在实际应用中具有重要价值。

关键词: 干扰抑制, 复数卷积神经网络, 间歇采样转发干扰, 合成孔径雷达图像处理, 深层特征

Abstract:

To address the problem that interrupted sampling repeater jamming （ISRJ） can severely impact the reliability of synthetic aperture radar （SAR） images, a suppression method of SAR ISRJ based on complex-valued deep learning using SAR image pairs with ISRJ, and a complex-valued deep residual shrinkage U-Net （CV-DRSUNet） are proposed. A complex-valued residual shrinkage block is constructed to explore the representation of target scenes and jamming signals in SAR complex-valued images, achieving ISRJ suppression. Additionally, a training strategy that does not rely on jamming-free scene images is proposed, which partially solves the issue of obtaining clean images in the presence of jamming in scenes, expanding the application of deep learning in jamming suppression. The effectiveness of the strategy and network model in ISRJ suppression is validated through mathematical derivation and experimental results. The proposed method exhibits important value in practical application.

Key words: jamming suppression, complex-valued convolutional neural network, interrupted sampling repeater jamming （ISRJ）, synthetic aperture radar （SAR） image processing, deep feature

中图分类号:

TN 974

陈兴宏, 张劲东, 万程. 基于含干扰数据的SAR图像域深度学习ISRJ抑制方法[J]. 系统工程与电子技术, 2026, 48(10): 3404-3416.

Xinghong Chen, Jindong Zhang, Cheng Wan. Deep learning method for ISRJ suppression in SAR image domain based on jammed data[J]. Systems Engineering and Electronics, 2026, 48(10): 3404-3416.

图/表 17

图1

图2

图3

图4

图5

图6

表1

图7

表2

表3

图8

图9

表4

表5

图10

表6

表7

参考文献 31

1	Fennell M T, Wishner R P. Battlefield awareness via synergistic SAR and MTI exploitation[J]. IEEE Aerospace and Electronic Systems Magazine, 1998, 13 (2): 39. doi: 10.1109/62.656334
2	Solimene R, Catapano I, Gennarelli G, et al. SAR imaging algorithms and some unconventional applications: a unified mathematical overview[J]. IEEE Signal Processing Magazine, 2014, 31 (4): 90. doi: 10.1109/MSP.2014.2311271
3	Wei S J, Zeng X F, Zhang H, et al. LFG-Net: low-level feature guided network for precise ship instance segmentation in SAR images[J]. IEEE Trans. on Geoscience and Remote Sensing, 2022, 60, 5231017.
4	Louhisuo M, Rauste Y, Andersson K, et al. Use of SAR data for natural disaster mitigation in the mobile environment[C]//Envisat and ERS Symposium, 2004: 6.
5	Oberstadler R, Hönsch H, Huth D. Assessment of the mapping capabilities of ERS-1 SAR data for flood mapping: a case study in Germany[J]. Hydrological Processes, 1997, 11 (10): 1415. doi: 10.1002/(SICI)1099-1085(199708)11:10<1415::AID-HYP532>3.0.CO;2-2
6	Gong S X, Wei X Z, Li X. ECCM scheme against interrupted sampling repeater jammer based on time-frequency analysis[J]. Journal of Systems Engineering and Electronics, 2014, 25 (6): 996. doi: 10.1109/JSEE.2014.00114
7	Zhou C, Liu Q H, Chen X L. Parameter estimation and suppression for DRFM-based interrupted sampling repeater jammer[J]. IET Radar, Sonar & Navigation, 2018, 12 (1): 56.
8	杨小鹏, 韩博文, 吴旭晨, 等. 基于短时分数阶傅里叶变换的间歇采样转发干扰辨识方法[J]. 信号处理, 2019, 35 (6): 1002.
9	Huan S, Dai G E, Luo G Y, et al. Bayesian compress sensing based countermeasure scheme against the interrupted sampling repeater jamming[J]. Sensors, 2019, 19 (15): 3279. doi: 10.3390/s19153279
10	Zhang Y, Wei Y S, Yu L. Interrupted sampling repeater jamming recognition and suppression based on phase-coded signal processing[J]. Signal Processing, 2022, 198, 108596. doi: 10.1016/j.sigpro.2022.108596
11	Wei S P, Zhang L, Lu J Y, et al. Modulated-ISRJ rejection using online dictionary learning for synthetic aperture radar imagery[J]. Journal of Systems Engineering and Electronics, 2023, 35 (2): 316.
12	Chen X Y, Chen B X. Interrupted-sampling repeater jamming suppression based on iterative decomposition[J]. Digital Signal Processing, 2023, 138, 104059. doi: 10.1016/j.dsp.2023.104059
13	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, 2015: 234.
14	Dong L H, Xu S, Xu B. Speech-transformer: a no-recurrence sequence-to-sequence model for speech recognition[C]//IEEE International Conference on Acoustics, Speech and Signal Processing, 2018: 5884.
15	Vaswani A, Shazeer N, Parmar N, et al. Attention is all you need[C]//31st International Conference on Neural Information Processing System, 2017: 6000.
16	Wei S J, Zhang H, Zeng X F, et al. CARNet: an effective method for SAR image interference suppression[J]. International Journal of Applied Earth Observation and Geoinformation, 2022, 114, 103019. doi: 10.1016/j.jag.2022.103019
17	Hua Q L, Zhang Y, Wei C X, et al. A self-supervised method based on CV-MUNet++ for active jamming suppression in SAR images[J]. IEEE Trans. on Geoscience and Remote Sensing, 2023, 61, 5209516.
18	Wu Y J, Duan L N, Yang L M, et al. Fine-grained recognition and suppression of ISRJ based on UNet-A[J]. IEEE Geoscience and Remote Sensing Letters, 2024, 21, 3448611.
19	Han B W, Qu X D, Yang X P, et al. DRFM-based repeater jamming reconstruction and cancellation method with accurate edge detection[J]. Remote Sensing, 2023, 15 (7): 1759. doi: 10.3390/rs15071759
20	Wei J Y, Li Y C, Yang R, et al. A time-domain filtering method based on intrapulse joint interpulse coding to counter interrupted sampling repeater jamming in SAR[J]. IEEE Trans. on Geoscience and Remote Sensing, 2023, 61, 5215912.
21	Wang X S, Liu J C, Zhang W M, et al. Mathematic principles of interrupted-sampling repeater jamming （ISRJ）[J]. Science in China Series F: Information Sciences, 2007, 50 (1): 113. doi: 10.1007/s11432-007-2017-y
22	Lehtinen J, Munkberg J, Hasselgren J, et al. Noise2Noise: learning image restoration without clean data[C]//35th International Conference on Machine Learning, 2018: 2965.
23	Zhang Z M, Wang H P, Xu F, et al. Complex-valued convolutional neural network and its application in polarimetric SAR image classification[J]. IEEE Trans. on Geoscience and Remote Sensing, 2017, 55 (12): 7177. doi: 10.1109/TGRS.2017.2743222
24	Yu L J, Hu Y H, Xie X C, et al. Complex-valued full convolutional neural network for SAR target classification[J]. IEEE Geoscience and Remote Sensing Letters, 2019, 17 (10): 1752.
25	Yu L J, Zeng Z X, Liu A, et al. A lightweight complex-valued DeepLabv3+ for semantic segmentation of PolSAR image[J]. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 2022, 15, 930. doi: 10.1109/JSTARS.2021.3140101
26	Zhu Y J, Li T, Peng D L, et al. A novel SAR automatic target recognition method based on fully complex-valued networks[J]. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 2023, 16, 6160. doi: 10.1109/JSTARS.2023.3292315
27	He K M, Zhang X Y, Ren S Q, et al. Deep residual learning for image recognition[C]//IEEE Conference on Computer Vision and Pattern Recognition, 2016: 770.
28	Zhao M H, Zhong S S, Fu X Y, et al. Deep residual shrinkage networks for fault diagnosis[J]. IEEE Trans. on Industrial Informatics, 2019, 16 (7): 4681.
29	Chen L C, Papandreou G, Kokkinos I, et al. Deeplab: semantic image segmentation with deep convolutional nets, Atrous convolution, and fully connected CRFs[J]. IEEE Trans. on Pattern Analysis and Machine Intelligence, 2017, 40 (4): 834.
30	Chen J, Wu W Z, Xu S Y, et al. Band pass filter design against interrupted-sampling repeater jamming based on time-frequency analysis[J]. IET Radar, Sonar & Navigation, 2019, 13 (10): 1646.
31	孙正阳, 董玫, 陈伯孝. 时频分析联合带通滤波抑制间歇采样转发干扰[J]. 西安电子科技大学学报, 2021, 48 (2): 139.

层数	下采样路径	上采样路径
0	64通道复数卷积层复数激活函数	单通道复数卷积层64通道复数卷积层复数激活函数
1	64通道CVRSB	64通道CVRSB
2	128通道CVRSB	128通道CVRSB
3	256通道CVRSB	256通道CVRSB
4	512通道CVRSB	512通道CVRSB
5	1024通道CVRSB	1024通道CVRSB
6	1024通道CVRSB	1024通道CVRSB

组别	距离向		方位向
组别	占空比	采样周期	占空比	采样周期
1	0.1	$T_p/5 $	0.05	$T_s/10 $
2	0.2	$T_p/5 $	0.1	$T_s/30 $
3	0.2	$T_p/10 $	0.075	$T_s/20 $

区域	参数	值
休斯顿	f₀/GHz	5.4017
	B/MHz	87.71
	Tp/μs	45.45
	Fs/MHz	100.0926
	脉冲重复频率/Hz	1870.6565
塞马塔岛	Fc/GHz	5.401
	B/MHz	59.44
	Tp/μs	44.172
	Fs/MHz	66.7284
	脉冲重复频率/Hz	1 925

指标	频域滤波	文献[31]	文献[9]	US-CV-ResUNet	US-CV-DRSUNet	CV-DRSUNet
SJRIF/dB	19.7777	11.7202	20.7871	23.6901	24.9724	25.4918
SSIM	0.9899	0.9638	0.9734	0.9954	0.9968	0.9972
PSNR/dB	52.6766	46.0007	52.6159	56.8495	57.5541	58.1777
IESR/dB	8.6195	5.5585	8.2313	8.7613	8.4759	8.5619

指标	频域滤波	文献[31]	文献[9]	US-CV-ResUNet	US-CV-DRSUNet	CV-DRSUNet
SJRIF/dB	21.1257	10.4741	14.9009	25.4148	26.6988	27.1672
SSIM	0.9866	0.9748	0.9596	0.9953	0.9966	0.9968
PSNR/dB	50.3242	45.9856	46.7760	55.5870	56.3161	56.8422
IESR/dB	6.0011	3.5362	7.5346	5.5543	5.3136	5.4013

基于含干扰数据的SAR图像域深度学习ISRJ抑制方法

Deep learning method for ISRJ suppression in SAR image domain based on jammed data

RichHTML

PDF (PC)

可视化

摘要/Abstract

引用本文

使用本文

图/表 17

参考文献 31

相关文章 15

编辑推荐

Metrics

本文评价

指标	US-RV- ResUNet	US-RV- DRSUNet	US-CV- ResUNet	US-CV- DRSUNet
SJRIF	13.7637	14.6352	22.3980	23.0361
PSNR	44.4287	44.4769	52.3618	53.5452
SSIM	0.9739	0.9736	0.9911	0.9947
IESR	4.7983	4.8521	5.0837	4.8180

测试方法	运算时间
文献[31]	11.1546
文献[9]	10.7006
CV-ResUNet	1.4864
CV-DRSUNet	1.7778

[1]	王思宇, 宋志勇, 王占领. 一种核密度估计与多域特征相结合的组合ISRJ-SMSP干扰识别方法[J]. 系统工程与电子技术, 2026, 48(6): 1838-1847.
[2]	罗颖聪, 张磊. 抗间歇采样转发干扰的分段失配滤波方法[J]. 系统工程与电子技术, 2026, 48(3): 795-806.
[3]	王陶然, 王福来, 祝迪, 邢世其, 殷加鹏, 黄晓涛, 李永祯. 基于时频拉伸变换的间歇采样转发干扰方法[J]. 系统工程与电子技术, 2026, 48(10): 3340-3350.
[4]	钟琦, 魏绍仁, 贺志毅, 刘建欣. 基于时频域行聚类的抗间歇采样转发干扰方法[J]. 系统工程与电子技术, 2026, 48(1): 106-118.
[5]	董乔龙, 刘思琪, 张伟见, 张劲东. LFM-DPC抗间歇采样转发干扰波形集设计算法[J]. 系统工程与电子技术, 2025, 47(7): 2154-2164.
[6]	周凯, 李超, 李坤, 崔新风, 姚辉伟, 陈冬冬. 抗移频间歇采样转发干扰波形与非匹配滤波器设计方法[J]. 系统工程与电子技术, 2025, 47(7): 2194-2204.
[7]	巴晓辉, 温雯霏, 蔡伯根, 王剑, 姜维, 柴琳果. GNSS信号的线性调频干扰抑制方法[J]. 系统工程与电子技术, 2025, 47(7): 2339-2348.
[8]	史朝鹏, 吴灏, 王万田, 朱轩, 张嘉毫, 孟进, 李强. 一种抗间歇采样转发干扰的脉内自适应凹陷LFM波形设计方法[J]. 系统工程与电子技术, 2025, 47(4): 1115-1126.
[9]	安强, 叶春茂, 鲁耀兵, 张彦. 复杂场景下间歇采样转发干扰抑制算法[J]. 系统工程与电子技术, 2025, 47(2): 463-477.
[10]	温镇铭, 王国宏, 张亮, 于洪波. 基于非最优阶次FrFT的间歇采样转发干扰抑制方法[J]. 系统工程与电子技术, 2025, 47(11): 3635-3643.
[11]	朱虹宇, 胡卫东, 王超, 施庆展, 张曦蒙, 袁乃昌. 基于神经网络的全双工干扰机自干扰对消方法[J]. 系统工程与电子技术, 2025, 47(1): 12-21.
[12]	刘柳, 刘田, 李洋, 潘文生, 张琰. 相控阵组网测控同频中继射频自干扰抑制技术[J]. 系统工程与电子技术, 2025, 47(1): 307-315.
[13]	潘小义, 刘晓斌, 陈吉源, 冯学文, 顾赵宇, 肖顺平. 间歇采样转发干扰技术研究述评[J]. 系统工程与电子技术, 2024, 46(9): 2887-2901.
[14]	于雷, 刘一品, 位寅生. 基于信干噪比最大化的盲提取抗主瓣干扰方法[J]. 系统工程与电子技术, 2024, 46(9): 2968-2979.
[15]	张明兴, 时巧, 余涛, 周正春, 崔国龙. 抗速度欺骗干扰的雷达发射波形与接收滤波器联合设计方法[J]. 系统工程与电子技术, 2024, 46(8): 2581-2591.