基于行为识别网络的SAR工作模式识别方法

doi:10.12305/j.issn.1001-506X.2026.04.12

系统工程与电子技术 ›› 2026, Vol. 48 ›› Issue (4): 1218-1229.doi: 10.12305/j.issn.1001-506X.2026.04.12

• 传感器与信号处理 • 上一篇

基于行为识别网络的SAR工作模式识别方法

胡继军¹, 韩伟¹^,*(), 张国玉¹, 李洋¹, 杨洁¹, 田甜²

1. 北京遥测技术研究所，北京 100076
2. 西安电子科技大学电子工程学院，陕西西安 710071

收稿日期:2025-02-06 修回日期:2025-04-03 接受日期:2025-07-18 出版日期:2025-05-23 发布日期:2025-05-23
通讯作者: 韩伟 E-mail:hanwei11111@126.com
作者简介:胡继军（1981—），男，研究员，博士研究生，主要研究方向为信息对抗、电子侦察和通信侦察
张国玉（1987—），男，高级工程师，硕士，主要研究方向为信息对抗和电子侦察
李　洋（1996—），男，工程师，硕士，主要研究方向为信息对抗、深度学习和人工智能
杨　洁（1981—），女，工程师，硕士，主要研究方向为信息对抗和电子侦察
田　甜（1992—），女，副教授，博士，主要研究方向为电子侦察、通信侦察和人工智能

SAR operating mode recognition method based on behavioral recognition network

Jijun HU¹, Wei HAN¹^,*(), Guoyu ZHANG¹, Yang LI¹, Jie YANG¹, Tian TIAN²

1. Beijing Remote Sensing Technology Research Institute，Beijing 100076，China
2. School of Electronic Engineering，Xidian University，Xi’an 710071，China

Received:2025-02-06 Revised:2025-04-03 Accepted:2025-07-18 Online:2025-05-23 Published:2025-05-23
Contact: Wei HAN E-mail:hanwei11111@126.com

摘要/Abstract

摘要：

针对星载合成孔径雷达工作模式识别准确率低及工程化难度高的问题，提出基于双重注意力网络-空洞卷积-Transformer行为网络的识别方法。所提方法结合Transformer网络在时间序列处理以及空洞卷积在扩大感受野和降低计算量上的优势，构建了以Transformer为骨架的工作模式识别网络，有效捕捉合成孔径雷达数据中潜在的时序特性。同时，为进一步增强模型对关键信息的敏感度，引入双重注意力网络，显著提高模型的识别准确率。仿真结果证明方法具有有效性，可以为遥感卫星的功能识别提供技术支持。

关键词: 星载合成孔径雷达, 工作模式识别, Transformer, 双重注意力网络, 空洞卷积

Abstract:

To address the issues of low recognition accuracy and difficulty in engineering implementation for spaceborne synthetic aperture radar （SAR） working mode identification, a method based on the dual attention network-dialated convolution-Transformer behavior network is proposed. It combines the advantages of the Transformer network in time series processing and dilated convolution in expanding the receptive field and reducing computational load, constructing a working mode recognition network with Transformer as the backbone to effectively capture the potential temporal characteristics in SAR data. Additionally, to further enhance the model’s sensitivity to key information, the dual attention network is introduced, significantly improving the model’s recognition accuracy. Simulation results demonstrate the effectiveness of the method, providing technical support for the functional identification of remote sensing satellites.

Key words: spaceborne synthetic aperture radar, operating mode identification, Transformer, dual attention network, dilated convolution

中图分类号:

胡继军, 韩伟, 张国玉, 李洋, 杨洁, 田甜. 基于行为识别网络的SAR工作模式识别方法[J]. 系统工程与电子技术, 2026, 48(4): 1218-1229.

Jijun HU, Wei HAN, Guoyu ZHANG, Yang LI, Jie YANG, Tian TIAN. SAR operating mode recognition method based on behavioral recognition network[J]. Systems Engineering and Electronics, 2026, 48(4): 1218-1229.

图/表 12

图1

图2

图3

表1

表2

图4

图5

图6

图7

表3

表4

图8

参考文献 30

1	LV J M, ZHU D Y, GENG Z H, et al. Recognition for SAR deformation military target from a new miniSAR dataset using multi-view joint transformer approach[J]. ISPRS Journal of Photogrammetry and Remote Sensing, 2024, 210 (1): 180- 197.
2	BERENS P, WALTERSCHEID I, SAALMANN O, et al. High resolution multi-aspect SAR imaging of military vehicles[C]// Proc. of the 13th European Conference on Synthetic Aperture Radar, 2021.
3	AN S H, KIM D J. SAR simulation of phase and amplitude images enhancing military target identification[C]// Proc. of the IEEE International Geoscience and Remote Sensing Symposium, 2023.
4	HU C, LI Y H, CHEN Z Y, et al. Distributed spaceborne SAR: a review of systems, applications, and the road ahead[J]. IEEE Geoscience and Remote Sensing Magazine, 2025, 13 (2): 329- 361.
5	SHEN Y, WANG X M, DAI K R, et al. Inference of creep landslide slip surface by InSAR technology and improved particle swarm optimization[J]. Landslides, 2025, 22, 1665- 1676.
6	CHANG T, YI Y H, JIANG H R, et al. Unraveling the non-linear relationship between seasonal deformation and permafrost active layer thickness[J]. NPJ Climate and Atmospheric Science, 2024, 7 (1): 308.
7	ZHOU X B, CHANG N B, LI S S. Applications of SAR interferometry in earth and environmental science research[J]. Sensors, 2009, 9 (3): 1876- 1912. doi: 10.3390/s90301876
8	SHAURY A, SAVANTH A S. Artificial neural network and fuzzy classifier for synthetic aperture radar image[C]// Proc. of the 2nd International Conference on Networks, Multimedia and Information Technology, 2024.
9	马志毅, 栗华美, 张雪峰, 等. 某星载降水雷达结构动力学优化设计方法[J]. 遥测遥控, 2024, 45 (4): 124- 131. doi: 10.12347/j.ycyk.20231221002
	MA Z Y, LI H M, ZHANG X F, et al. Structural dynamics optimization design method for a spaceborne precipitation radar[J]. Journal of Telemetry, Tracking and Command, 2024, 45 (4): 124- 131. doi: 10.12347/j.ycyk.20231221002
10	MACCHIARULO V, GIARDINA G, MILILLO P, et al. Integrating post-event very high resolution SAR imagery and machine learning for building-level earthquake damage assessment[J]. Bulletin of Earthquake Engineering, 2024, 22 (2): 345- 360.
11	RODGER M, GUIDA R. Damage assessment mapping in Mariupol （Ukraine） with multi-temporal synthetic aperture radar （SAR）[J]. IEEE Trans. on Geoscience and Remote Sensing, 2023, 61 (9): 7186- 7189.
12	WANG W D, MOTAGH M, XIA Z G, et al. A framework for automated landslide dating utilizing SAR-derived parameters time-series, an enhanced transformer model, and dynamic thresholding[J]. International Journal of Applied Earth Observation and Geoinformation, 2024, 129 (1): 103795.
13	TAO M L, SU J, HUANG Y, et al. Mitigation of radio frequency interference in synthetic aperture radar data: current status and future trends[J]. Remote Sensing, 2019, 11 (20): 2438. doi: 10.3390/rs11202438
14	HUANG B, WANG W Q, ZHANG S S H, et al. A novel approach for spaceborne SAR scattered-wave deception jamming using frequency diverse array[J]. IEEE Geoscience and Remote Sensing Letters, 2020, 17(9): 1568-1572.
15	WAQAS M, HUMPHRIES U W. A critical review of RNN and LSTM variants in hydrological time series predictions[J]. MethodsX, 2024, 13, 102946. doi: 10.1016/j.mex.2024.102946
16	KOUCHAKI M, ZHANG M L, ABDALLA A S, et al. Enhanced real-time threat detection in 5G networks: a self-attention RNN autoencoder approach for spectral intrusion analysis[C]// Proc. of the 22nd International Symposium on Modeling and Optimization in Mobile, Ad Hoc, and Wireless Networks, 2024: 249−256.
17	TAMILSELVI C, PAUL R K, YEASIN M, et al. Novel wavelet-LSTM approach for time series prediction[J]. Neural Computing and Applications, 2024, 36 (5): 10521- 10530.
18	CLERICO V, GONZALEZ-LOPEZ J, GADY AGAM, et al. LSTM framework for classification of radar and communications signals[C]// Proc. of the IEEE Radar Conference, 2023.
19	VASWANI A, SHAZEER N, PARMAR N, et al. Attention is all you need[J]. Advances in Neural Information Processing Systems, 2017, 30, 5998- 6008.
20	CHEN H Y, FENG K G, KONG Y K, et al. Multi-function radar work mode recognition based on encoder-decoder model[C]// Proc. of the IEEE International Geoscience and Remote Sensing Symposium, 2022: 1189−1192.
21	MA R F, JIN G D, SONG C, et al. A novel method to identify the spaceborne SAR operating mode based on delobe reconnaissance and machine learning[J]. Remote Sensing, 2024, 16 (7): 1234. doi: 10.3390/rs16071234
22	CHI K, SHEN J H, LI Y, et al. A novel segmentation approach for work mode boundary detection in MFR pulse sequence[J]. Digital Signal Processing, 2022, 126, 103462. doi: 10.1016/j.dsp.2022.103462
23	XIONG J W, PAN J F, DU M Y. A cascade network for pattern recognition based on radar signal characteristics in noisy environments[J]. Remote Sensing, 2023, 15 (8): 2045. doi: 10.3390/rs15082045
24	DU M Y, ZHAO P, CAI X H, et al. Robust bayesian attention belief network for radar work mode recognition[J]. Digital Signal Processing, 2023, 133: 103874.
25	KRIZHEVSKY A, SUTSKEVER I, HINTON G. ImageNet classification with deep convolutional neural networks[J]. Communications of the ACM, 2017, 60 (6): 1097- 1105.
26	DONG X X, CHENG S Y. Radar working modes recognition based on discrete process neural network[C]// Proc. of the IOP Conference Series: Materials Science and Engineering, 2018, 394: 042088.
27	ZHANG Y J, ZHANG C, HUO W B. Operation mode recognition of airborne radar based on multi-feature fusion RS-ConvNet[C]// Proc. of the IEEE International Geoscience and Remote Sensing Symposium, 2022.
28	XIONG R S, YANG Y C, HE D, et al. On layer normalization in the transformer architecture [C]// Proc. of the 37th International Conference on Machine Learning, 2020, 119: 10524−10533.
29	JIAO J Y, ZHANG Y, ZHANG Y, et al. Dilateformer: multi-scale dilated transformer for visual recognition[J]. IEEE Trans. on Multimedia, 2023, 25, 8906- 8919. doi: 10.1109/TMM.2023.3243616
30	FU J, LIU J, TIAN H J, et al. Dual attention network for scene segmentation[C]// Proc. of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2020: 3140−3149.

工作模式	载频/GHz	脉宽/μs	带宽/MHz	幅度/dB	重复间隔/μs
条带	5.4	24.3	130~190	−66~–48	202.6
聚束	5.4	25.05	270~330	−50	208.75
滑动聚束	5.4	23	10~70	−55~−48	190
扫描	5.4	23.5/23.25/23.1	10~70	−66~−49	196/193.8/192.3

参数名称	配置
训练样本数	1600
样本类别	4
优化方法	Adam
批大小	128
样本维度	[128,5]
学习率	5e-4
训练轮数	30

网络模块	Transformer	空洞卷积	DANet	Accuracy/%
1	√	×	×	79.47
2	√	×	√	86.02
3	√	√	×	94.87
4	√	√	√	99.12

指标类别	网络结构
指标类别	CNN	CGRU	Informer	所提网络
准确率/%	85.96	91.85	95.22	99.12
参数数量（×10⁶）	0.148	0.229	0.162	0.154
计算复杂度（×10⁶）	3.687	9.168	5.086	4.870

[1]	李鑫晟, 冯书谊, 郝禹哲, 叶曦, 张海超, 李元祥. 基于多源卫星遥感数据的舰船航迹关联与融合[J]. 系统工程与电子技术, 2026, 48(4): 1154-1164.
[2]	何肖阳, 陈小龙, 杜晓林, 苏宁远, 袁旺, 关键. 基于CBAM-Swin-Transformer迁移学习的海上微动目标分类方法[J]. 系统工程与电子技术, 2025, 47(4): 1155-1167.
[3]	邵凯, 吴广, 梁燕, 奚兴发, 高琳珈. 基于局部特征编解码的自动驾驶3D目标检测[J]. 系统工程与电子技术, 2025, 47(10): 3168-3178.
[4]	阳鹏飞, 何羚, 王茜, 王睿笛, 张明志. 基于混合信号多域特征和Transformer的干扰识别[J]. 系统工程与电子技术, 2024, 46(6): 2138-2145.
[5]	刘子昌, 白永生, 李思雨, 贾希胜. 基于小波时频图与Swin Transformer的柴油机故障诊断方法[J]. 系统工程与电子技术, 2023, 45(9): 2986-2998.
[6]	眭海刚, 李嘉杰, 苟国华. 基于异源影像匹配的无人机在线快速定位方法[J]. 系统工程与电子技术, 2023, 45(10): 3008-3015.
[7]	钱坤, 李晨瑄, 陈美杉, 郭继伟, 潘磊. 基于改进Swin Transformer的舰船目标实例分割算法[J]. 系统工程与电子技术, 2023, 45(10): 3049-3057.
[8]	魏文晓, 刘洁瑜, 沈强, 李成. 基于人眼视点图的特征融合小目标检测算法[J]. 系统工程与电子技术, 2022, 44(4): 1120-1127.
[9]	苏娟, 杨龙, 黄华, 金国栋. 用于SAR图像小目标舰船检测的改进SSD算法[J]. 系统工程与电子技术, 2020, 42(5): 1026-1034.
[10]	周超伟, 李真芳, 毛琴, 张金强, 索志勇. 高分辨率滑动聚束SAR卫星姿态机动策略[J]. 系统工程与电子技术, 2018, 40(7): 1472-1477.
[11]	张延, 郭苹, 张林让, 唐世阳. 高分辨星载SAR的方位维频率非线性变标成像算法[J]. 系统工程与电子技术, 2017, 39(8): 1716-1722.
[12]	马喜乐, 孙造宇, 董臻. 优化DPC-MAB SAR性能的自适应阵列配置技术[J]. Journal of Systems Engineering and Electronics, 2011, 33(11): 2409-2412.
[13]	杨光，李绍滨，姜义成. 基于分数阶傅里叶变换的星载SAR信号参数实时估计[J]. Journal of Systems Engineering and Electronics, 2010, 32(8): 1649-1651.
[14]	魏雪云, 李景文, 徐华平. 星载合成孔径雷达影像几何校正[J]. Journal of Systems Engineering and Electronics, 2010, 32(7): 1422-1425.

基于行为识别网络的SAR工作模式识别方法

SAR operating mode recognition method based on behavioral recognition network

RichHTML

PDF (PC)

可视化

摘要/Abstract

引用本文

使用本文

图/表 12

参考文献 30

相关文章 14

编辑推荐

Metrics

本文评价