DDIRNet: robust radar emitter recognition via single domain generalization

doi:10.23919/JSEE.2025.000053

Journal of Systems Engineering and Electronics ›› 2025, Vol. 36 ›› Issue (2): 397-404.doi: 10.23919/JSEE.2025.000053

• • 上一篇

收稿日期:2023-08-07 接受日期:2023-08-07 出版日期:2025-04-18 发布日期:2025-05-20

DDIRNet: robust radar emitter recognition via single domain generalization

Honglin WU(), Xueqiong LI(), Junjie HUANG(), Ruochun JIN(), Yuhua TANG()

Institute for Quantum Information & State Key Laboratory of High Performance Computing, National University of Defense Technology, Changsha 410073, China

Received:2023-08-07 Accepted:2023-08-07 Online:2025-04-18 Published:2025-05-20
Contact: Xueqiong LI E-mail:honglinwu@nudt.edu.cn;lixueqiong13@nudt.edu.cn;jjhuang@nudt.edu.cn;jinrc@nudt.edu.cn;yhtang62@163.com
About author:
WU Honglin was born in 1999. He received his M.S. degree from the College of Computer Science and Technology, National University of Defense Technology, Changsha, China, in 2021. He is working toward his Ph.D. degree in computer science and technology from National University of Defense Technology. His research interests include signal processing and deep learning. E-mail: honglinwu@nudt.edu.cn

LI Xueqiong was born in 1991. She received her Ph.D. degree from the College of Computer Science and Technology, National University of Defense Technology, Changsha, China, in 2020. She is currently an associate researcher with the College of Computer Science and Technology, National University of Defense Technology. Her current research interests include signal processing and machine learning. E-mail: lixueqiong13@nudt.edu.cn

HUANG Junjie was born in 1990. He received his Ph.D. degree from Imperial College London, London, U.K., in 2019. He is currently a lecturer with the College of Computer Science and Technology, National University of Defence Technology, Changsha, China. During 2019–2021, he was a Postdoc with Communications and Signal Processing Group, Electrical and Electronic Engineering Department, ICL, London. His current research interests include explainable deep learning and computer vision. E-mail: jjhuang@nudt.edu.cn

JIN Ruochun was born in 1993. He received his Ph.D. degree from the College of Computer Science and Technology, National University of Defense Technology, Changsha, China, in 2021. He is currently an assistant researcher with the College of Computer Science and Technology, National University of Defense Technology. His current research interests include databases, data quality, and data science. E-mail: jinrc@nudt.edu.cn

TANG Yuhua was born in 1962. She received her M.S degree from College of Computer Scienceand Technology, National University of Defense Technology, Changsha, China, in 1986. She is currently a professor with the Institute for Quantum Information & State Key Laboratory of High Performance Computing, College of Computer Science and Technology, National University of Defense Technology.Her research interests include supercomputer architecture and core router design E-mail: yhtang62@163.com
Supported by:
This work was supported by the National Natural Science Foundation of China (62101575), the Research Project of NUDT (ZK22-57), and the Self-directed Project of State Key Laboratory of High Performance Computing (202101-16).

摘要/Abstract

Abstract:

Automatically recognizing radar emitters from complex electromagnetic environments is important but non-trivial. Moreover, the changing electromagnetic environment results in inconsistent signal distribution in the real world, which makes the existing approaches perform poorly for recognition tasks in different scenes. In this paper, we propose a domain generalization framework is proposed to improve the adaptability of radar emitter signal recognition in changing environments. Specifically, we propose an end-to-end denoising based domain-invariant radar emitter recognition network (DDIRNet) consisting of a denoising model and a domain invariant representation learning model (IRLM), which mutually benefit from each other. For the signal denoising model, a loss function is proposed to match the feature of the radar signals and guarantee the effectiveness of the model. For the domain invariant representation learning model, contrastive learning is introduced to learn the cross-domain feature by aligning the source and unseen domain distribution. Moreover, we design a data augmentation method that improves the diversity of signal data for training. Extensive experiments on classification have shown that DDIRNet achieves up to 6.4% improvement compared with the state-of-the-art radar emitter recognition methods. The proposed method provides a promising direction to solve the radar emitter signal recognition problem.

Key words: radar emitter recognition, domain generalization, denoising, contrastive learning, data augmentation

. [J]. Journal of Systems Engineering and Electronics, 2025, 36(2): 397-404.

Honglin WU, Xueqiong LI, Junjie HUANG, Ruochun JIN, Yuhua TANG. DDIRNet: robust radar emitter recognition via single domain generalization[J]. Journal of Systems Engineering and Electronics, 2025, 36(2): 397-404.

图/表 10

参考文献 30

1	WILEY R. ELINT: the interception and analysis of radar signals. Boston: Artech, 2006.
2	KEN’ICHI N, MASAAKI K Improved algorithm for estimating pulse repetition intervals. IEEE Trans. on Aerospace and Electronic Systems, 2000, 36 (2): 407- 421. doi: 10.1109/7.845217
3	SONG K H, LEE D W, HAN J W, et al Pulse repetition interval modulation recognition using symbolization. Proc. of the International Conference on Digital Image Computing: Techniques and Applications, 2010, 540- 545.
4	WILLSON G B Radar classification using a neural network. Proc. of the Defense, Security and Sensing, 1990, 200- 210.
5	FAN T, YU X X, GAN N, et al Transmit-receive design for airborne radar with nonuniform pulse repetition intervals. IEEE Trans. on Aerospace and Electronic Systems, 2021, 57 (6): 4067- 4084. doi: 10.1109/TAES.2021.3090915
6	LI X Q, HUANG Z T, WANG F, et al Toward convolutional neural networks on pulse repetition interval modulation recognition. IEEE Communications Letters, 2018, 22 (11): 2286- 2289. doi: 10.1109/LCOMM.2018.2864725
7	LI X Q, LIU Z M, HUANG Z T Attention-based radar primodulation recognition with recurrent neural networks. IEEE Access, 2020, 8, 57 426- 57 436. doi: 10.1109/ACCESS.2020.2982654
8	QU Q Z, WEI S J, WU Y, et al ACSE networks and autocorrelation features for pri modulation recognition. IEEE Communications Letters, 2020, 24 (8): 1729- 1733, 2020. doi: 10.1109/LCOMM.2020.2992266
9	DADGARNIA A, SADEGHI M T Automatic recognition of pulse repetition interval modulation using temporal convolutional network. IET Signal Processing, 2021, 15 (9): 633- 648. doi: 10.1049/sil2.12069
10	FENG H C, TANG B, WAN T Radar pulse repetition interval modulation recognition with combined net and domain-adaptive few-shot learning. Digital Signal Processing, 2022, 127, 103562. doi: 10.1016/j.dsp.2022.103562
11	PENG X B, ANDRYCHOWICZ M, ZAREMBA W, et al Sim-to-real transfer of robotic control with dynamics randomization. Proc. of the IEEE International Conference on Robotics and Automation, 2018, 3803- 3810.
12	KHIRODKAR R, YOO D, KITANI K Domain randomization for scene-specific car detection and pose estimation. Proc. of the IEEE Winter Conference on Applications of Computer Vision, 2019, 193- 1940.
13	BORLINO F, D’INNOCENTE A, TOMMASI T Domain generalization vs data augmentation: an unbiased perspective. Proc. of the European Conference on Computer Vision Workshops, 2020, : 726- 730.
14	LI P, LI D, LI W, et al A simple feature augmentation for domain generalization. Proc. of the IEEE/CVF International Conference on Computer Vision, 2021, 8866- 8875.
15	CHEN T L, BAKTASHMOTLAGH M, WANG Z J, et al Center-aware adversarial augmentation for single domain generalization. Proc. of the IEEE/CVF Winter Conference on Applications of Computer Vision, 2023, 4146- 4154.
16	XU X, ZHOU X, VENKATESAN R, et al D-SNE: domain adaptation using stochastic neighborhood embedding. Proc. of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2019, 2497- 2506.
17	QIAO F C, ZHAO L, PENG X Learning to learn single domain generalization. Proc. of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2020, 12556- 12565.
18	WAN C Q, SHEN X, ZHANG Y G, et al Meta convolutional neural networks for single domain generalization. Proc. of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2022, 4672- 4681.
19	CHEN J, GAO Z, WU X X, et al Meta-causal learning for single domain generalization. Proc. of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2023, 7683- 7692.
20	KIM H, KANG Y, OH C, et al Single domain generalization for LiDAR semantic segmentation. Proc of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2023, 17587- 17598.
21	ZHU W, LU L, XIAO J, et al Localized adversarial domain generalization. Proc. of the IEEE Conference on Computer Vision and Pattern Recognition, 2022, 7098- 7108.
22	ARJOVSKY M, BOTTOU L, GULRA-JANI I, et al. Invariant risk minimization. https://arxiv.org/abs/1907.02893.
23	LI L, GAO K, CAO J, et al Progressive domain expansion network for single domain generalization. Proc. of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2021, 224- 233.
24	WANG J D, LAN C L, LIU C, et al Generalizing to unseen domains: a survey on domain generalization. IEEE Trans. on Knowledge and Data Engineering, 2022, 35 (8): 8052- 8072.
25	CHEN T, KORNBLITH S, NOROUZI M, et al A simple framework for contrastive learning of visual representations. Proc. of the International Conference on Machine Learning, 2020, 1597- 1607.
26	LEE K Prototypical contrastive predictive coding. Proc. of the Tenth International Conference on Learning Representations, 2022.
27	VASWANI A, SHAZEER N, PARMAR N, et al Attention is all you need. Proc. of the Annual Conference on Neural Information Processing Systems, 2017, 5998- 6008.
28	GHORBEL E, GHORBEL M, M'HIRI S, et al Data augmentation based on invariant shape blending for deep learning classification. Proc. of the IEEE International Conference on Acoustics, Speech and Signal Processing, 2023, 1- 5.
29	ALJUNDI R, PATEL Y, SULC M, et al Contrastive classification and representation learning with probabilistic interpretation. Proc. of the Conference on Artificial Intelligence, 2023, 6675- 6683.
30	CHEN S, GONG C, LI J, et al Learning contrastive embedding in low-dimensional space. Proc. of the NeurIPS, 2022.

Dataset	$ \rho_r $	$ \rho_m $	$ \rho_n $
$ P_{{\mathrm{train}}} $	0.05	0.2	0.4
$ P_1 $	0.02	0.05	0.2
$ P_2 $	0.05	0.2	0.4
$ P_3 $	0.05	0.3	0.6
$ P_4 $	0.1	0.5	0.8

Method	$ P_1 $	$ P_2 $	$ P_3 $	$ P_4 $	Average
A-RNN [7]	86.2	93.7	75.1	50.5	76.4
ACSE [8]	86.6	93.1	74.9	49.0	75.9
TCN [9]	87.0	94.2	74.6	52.2	77.0
CombinedNet [10]	89.7	94.2	80.4	60.5	81.2
d-SNE [17]	88.9	93.1	78.6	59.7	80.1
MADA [18]	89.5	93.6	81.6	61.4	81.6
MetaCNN [19]	90.1	93.1	82.9	64.7	82.7
DDIRNet	92.9	93.9	89.1	74.5	87.6

Method	CST	JIT	SLD	WOB	D&S	STG
A-RNN [7]	40.4	40.6	59.3	57.7	53.0	50.8
ACSE [8]	40.0	38.6	53.4	58.6	54.5	49.0
TCN [9]	46.4	48.3	56.3	54.9	58.5	51.6
CombinedNet [10]	53.1	57.6	65.2	67.0	62.7	59.8
d-SNE [17]	58.5	56.1	64.6	62.0	59.2	59.2
MADA [18]	56.6	55.9	65.3	70.7	62.8	60.8
MetaCNN [19]	59.9	58.4	69.2	74.8	66.6	63.2
DDIRNet	71.0	61.7	80.5	84.4	83.1	72.9

Method	STG1	STG2	STG3	STG4	STG5
A-RNN [7]	52.8	51.1	45.1	49.1	56.1
ACSE [8]	50.4	53.2	49.5	47.9	45.1
TCN [9]	56.3	53.6	47.0	47.0	54.8
CombinedNet [10]	65.5	54.7	53.6	58.9	66.7
d-SNE [17]	63.5	61.7	53.2	56.2	62.8
MADA [18]	62.8	60.9	61.7	57.2	63.4
MetaCNN [19]	65.9	62.4	64.5	59.8	65.1
DDIRNet	78.3	75.1	66.8	69.9	74.4

$ D $	$ P $	$ P_1 $	$ P_2 $	$ P_3 $	$ P_4 $	Average
×	×	87.2	92.3	83.3	67.8	82.7
√	×	92.7	93.1	85.5	71.0	85.6
×	√	91.1	94.0	87.4	70.3	85.7
√	√	92.9	93.9	89.1	74.5	87.6

DDIRNet: robust radar emitter recognition via single domain generalization

RichHTML

PDF (PC)

可视化

摘要/Abstract

引用本文

使用本文

图/表 10

参考文献 30

相关文章 0

编辑推荐

Metrics

本文评价