Weak Labeling-specific Emitter Identification Algorithm Based on the Weakly Supervised Wav-KAN Network

LIU Kangsheng; LING Qing; YAN Wenjun; ZHANG Limin; YU Keyuan; LIU Hengyan

doi:10.12000/JR24248

Turn off MathJax

Article Contents

Abstract

References

Article Navigation > Journal of Radars > 2025 > Online First

LIU Kangsheng, LING Qing, YAN Wenjun, et al. Weak labeling-specific emitter identification algorithm based on the weakly supervised Wav-KAN network[J]. Journal of Radars, 2025, 14(2): 338–352. doi: 10.12000/JR24248

Citation:

LIU Kangsheng, LING Qing, YAN Wenjun, et al. Weak labeling-specific emitter identification algorithm based on the weakly supervised Wav-KAN network[J]. Journal of Radars, 2025, 14(2): 338–352. doi: 10.12000/JR24248

Citation:

PDF( 1833 KB)

Weak Labeling-specific Emitter Identification Algorithm Based on the Weakly Supervised Wav-KAN Network

DOI: 10.12000/JR24248 CSTR: 32380.14.JR24248

Institute of Information Fusion, Naval Aviation University, Yantai 264001, China

Funds: The National Natural Science Foundation of China (62371465), The Taishan Scholars Project Special Fund (ts201511020), Youth Innovation Teams in Shandong Province Fund (2022KJ084)

More Information

Corresponding author: LING Qing, linqing19870522@163.com; YAN Wenjun, wj_yan@foxmail.com
Received Date: 2024-12-11
Rev Recd Date: 2025-03-04

Available Online: 2025-03-12

Abstract

Abstract

Most existing specific emitter identification technologies rely on supervised learning, making them unsuitable for scenarios with label loss due to factors such as the acquisition environment (e.g., weather conditions, terrain, obstacles, and interference sources), device performance (e.g., radar resolution, signal processing capabilities, and hardware failures), and tagger level. In this study, a weakly labeled specific emitter identification algorithm based on the Weakly Supervised Wav-KAN (WSW-KAN) network is proposed. First, a WSW-KAN baseline network is constructed by integrating the unique learnable edge function of the KAN network with the multiresolution analysis of the wavelet function. The weakly labeled dataset is then divided into a small labeled dataset and a large unlabeled dataset, with the small labeled dataset used for initial model training. Finally, based on the pretrained model, Adaptive Pseudo-Label Weighted Selection (APLWS) is used to extract features from the unlabeled data using a contrast learning method, followed by iterative training, thereby effectively improving the generalization capability of the model. Experimental validation using a real acquisition radar dataset demonstrates that the proposed algorithm achieves a recognition accuracy of approximately 95% for specific emitters while maintaining high efficiency, a small parameter scale, and strong adaptability, making it suitable for practical applications.
- Specific Emitter Identification (SEI),
- Weakly Supervised Wav-KAN (WSW-KAN),
- Pseudo-label iteration,
- Weakly supervised learning,
- Contrastive learning,
- Neural network

FullText(HTML)

References(27)

References

[1]	吴琼. 基于改进CNN的雷达辐射源识别算法研究[D]. [硕士论文], 西安电子科技大学, 2019: 1–4. doi: 10.7666/d.D01905147. WU Qiong. Research on radar emitter recognition algorithm based on improved CNN[D]. [Master dissertation], Xidian University, 2019: 1–4. doi: 10.7666/d.D01905147.
[2]	TAN Kaiwen, YAN Wenjun, ZHANG Limin, et al. Specific emitter identification based on software-defined radio and decision fusion[J]. IEEE Access, 2021, 9: 86217–86229. doi: 10.1109/ACCESS.2021.3088542.
[3]	王艺卉, 闫文君, 凌青, 等. 基于置信学习的低标注率辐射源个体识别算法[J]. 兵器装备工程学报, 2024, 45(5): 267–275. doi: 10.11809/bqzbgcxb2024.05.037. WANG Yihui, YAN Wenjun, LING Qing, et al. Low labeling rate specific emitter identification algorithm based on confidence learning[J]. Journal of Ordnance Equipment Engineering, 2024, 45(5): 267–275. doi: 10.11809/bqzbgcxb2024.05.037.
[4]	张顺生, 丁宦城, 王文钦. 面向辐射源识别的多尺度特征提取与特征选择网络[J]. 国防科技大学学报, 2024, 46(6): 141–148. doi: 10.11887/j.cn.202406015. ZHANG Shunsheng, DING Huancheng, and WANG Wenqin. Multi-scale feature extraction and feature selection network for radiation source identification[J]. Journal of National University of Defense Technology, 2024, 46(6): 141–148. doi: 10.11887/j.cn.202406015.
[5]	XING Yuexiu, HU Aiqun, ZHANG Junqing, et al. Design of a robust radio-frequency fingerprint identification scheme for multimode LFM radar[J]. IEEE Internet of Things Journal, 2020, 7(10): 10581–10593. doi: 10.1109/JIOT.2020.3003692.
[6]	李昕. 基于无监督学习的通信辐射源个体识别技术研究[D]. [硕士论文], 国防科技大学, 2019: 1–10. doi: 10.27052/d.cnki.gzjgu.2019.000718. LI Xin. Research on individual communication transmitter identification based on unsupervised learning[D]. [Master dissertation], National University of Defense Technology, 2019: 1–10. doi: 10.27052/d.cnki.gzjgu.2019.000718.
[7]	CHEN Qi, YANG Lingxiao, LAI Jianhuang, et al. Self-supervised image-specific prototype exploration for weakly supervised semantic segmentation[C]. 2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), New Orleans, USA, 2022: 4278–4288. doi: 10.1109/CVPR52688.2022.00425.
[8]	DU Ye, FU Zehua, LIU Qingjie, et al. Weakly supervised semantic segmentation by pixel-to-prototype contrast[C]. 2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), New Orleans, USA, 2022: 4310–4319. doi: 10.1109/CVPR52688.2022.00428.
[9]	JIANG Pengtao, YANG Yuqi, HOU Qibin, et al. L2G: A simple local-to-global knowledge transfer framework for weakly supervised semantic segmentation[C]. 2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), New Orleans, USA, 2022: 16865–16875. doi: 10.1109/CVPR52688.2022.01638.
[10]	LEE J, KIM E, and YOON S. Anti-adversarially manipulated attributions for weakly and semi-supervised semantic segmentation[C]. 2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), Nashville, USA, 2021: 4070–4078. doi: 10.1109/CVPR46437.2021.00406.
[11]	LEE S, LEE M, LEE J, et al. Railroad is not a train: Saliency as pseudo-pixel supervision for weakly supervised semantic segmentation[C]. 2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), Nashville, USA, 2021: 5491–5501. doi: 10.1109/CVPR46437.2021.00545.
[12]	ZHOU Tianfei, ZHANG Meijie, ZHAO Fand, et al. Regional semantic contrast and aggregation for weakly supervised semantic segmentation[C]. 2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), New Orleans, USA, 2022: 4289–4299. doi: 10.1109/CVPR52688.2022.00426.
[13]	HE Kaiming, FAN Haoqi, WU Yuxin, et al. Momentum contrast for unsupervised visual representation learning[C]. 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), Seattle, USA, 2020: 9726–9735. doi: 10.1109/CVPR42600.2020.00975.
[14]	PENG Yang, HOU Changbo, ZHANG Yibin, et al. Supervised contrastive learning for RFF identification with limited samples[J]. IEEE Internet of Things Journal, 2023, 10(19): 17293–17306. doi: 10.1109/JIOT.2023.3272628.
[15]	CHEN Jun, WONG W K, and HAMDAOUI B. Unsupervised contrastive learning for robust RF device fingerprinting under time-domain shift[C]. ICC 2024 - IEEE International Conference on Communications, Denver, USA, 2024: 3567–3572. doi: 10.1109/ICC51166.2024.10622173.
[16]	ZHAN Mengqi, LI Yang, CUI Huajun, et al. MCRFF: A meta-contrastive learning-based RF fingerprinting method[C]. MILCOM 2023 - 2023 IEEE Military Communications Conference (MILCOM), Boston, USA, 2023: 391–396. doi: 10.1109/MILCOM58377.2023.10356211.
[17]	DU Mingyang, PAN Jifei, and BI Daping. A contrastive learner for automatic modulation classification[J]. IEEE Transactions on Wireless Communications, in press. doi: 10.1109/TWC.2025.3532438.
[18]	LIU Ziming, WANG Yixuan, VAIDYA S, et al. KAN: Kolmogorov-Arnold networks[EB/OL]. https://arxiv.org/abs/2404.19756, 2024.
[19]	BOZORGASL Z and CHEN Hao. Wav-KAN: Wavelet Kolmogorov-Arnold networks[EB/OL]. https://arxiv.org/abs/2405.12832, 2024.
[20]	PERSONS J B, WONG L J, MOORE M O, et al. Classification of radio signals using truncated Gaussian discriminant analysis of convolutional neural network-derived features[C]. MILCOM 2022 - 2022 IEEE Military Communications Conference (MILCOM), Rockville, USA, 2022: 304–310. doi: 10.1109/MILCOM55135.2022.10017724.
[21]	ZHENG Qinghe, ZHAO Penghui, LI Yang, et al. Spectrum interference-based two-level data augmentation method in deep learning for automatic modulation classification[J]. Neural Computing and Applications, 2021, 33(13): 7723–7745. doi: 10.1007/s00521-020-05514-1.
[22]	CHEN Xi, YANG Haosen, ZHANG Huicong, et al. Uncertainty-aware pseudo-label filtering for source-free unsupervised domain adaptation[J]. Neurocomputing, 2024, 575: 127190. doi: 10.1016/j.neucom.2023.127190.
[23]	刘宁波, 丁昊, 黄勇, 等. X波段雷达对海探测试验与数据获取年度进展[J]. 雷达学报, 2021, 10(1): 173–182. doi: 10.12000/JR21011. LIU Ningbo, DING Hao, HUANG Yong, et al. Annual progress of the sea-detecting X-band radar and data acquisition program[J]. Journal of Radars, 2021, 10(1): 173–182. doi: 10.12000/JR21011.
[24]	ALBATINEH A N, BOUBAKARI I, and KIBRIA B M G. New confidence interval estimator of the signal-to-noise ratio based on asymptotic sampling distribution[J]. Communications in Statistics - Theory and Methods, 2017, 46(2): 574–590. doi: 10.1080/03610926.2014.1000498.
[25]	ZAGORUYKO S and KOMODAKIS N. Wide residual networks[C]. British Machine Vision Conference 2016, York, UK, 2016.
[26]	谭凯文, 张立民, 闫文君, 等. 面向非均衡类别的半监督辐射源识别方法[J]. 雷达学报, 2022, 11(4): 713–727. doi: 10.12000/JR22043. TAN Kaiwen, ZHANG Limin, YAN Wenjun, et al. A semi-supervised emitter identification method for imbalanced category[J]. Journal of Radars, 2022, 11(4): 713–727. doi: 10.12000/JR22043.
[27]	王程昱, 凌青, 闫文君. 基于DCGAN的雷达辐射源信号个体识别算法[J]. 测控技术, 2024, 43(7): 17–22, 64. doi: 10.19708/j.ckjs.2024.04.222. WANG Chengyu, LING Qing, and YAN Wenjun. Individual identification algorithm for radar emitter signal based on DCGAN[J]. Measurement and Control Technology, 2024, 43(7): 17–22, 64. doi: 10.19708/j.ckjs.2024.04.222.