领域知识内嵌的电磁信号调制方式智能识别

赵宏佳; 张多纳; 鲁远耀; 丁文锐

doi:10.13700/j.bh.1001-5965.2023.0746

领域知识内嵌的电磁信号调制方式智能识别

doi: 10.13700/j.bh.1001-5965.2023.0746

1.
北方工业大学信息学院，北京 100144
2.
北京航空航天大学无人系统研究院，北京 100191

基金项目:

国家自然科学基金(62201010)；北京市教育委员会科研计划项目(KM202310009003)

详细信息

通讯作者:
E-mail：zhangduona@buaa.edu.cn

中图分类号: TN911
计量
- 文章访问数: 950
- HTML全文浏览量: 232
- PDF下载量: 67
- 被引次数: 0
出版历程
- 收稿日期: 2023-11-20
- 录用日期: 2024-03-11
- 网络出版日期: 2024-03-13
- 整期出版日期: 2026-01-31

Intelligent recognition of electromagnetic signal modulation with embedded domain knowledge

1.
School of Information Science and Technology，North China University of Technology，Beijing 100144，China
2.
Unmanned Systems Research Institute，Beihang University，Beijing 100191，China

Funds:

National Natural Science Foundation of China (62201010); R&D Program of Beijing Municipal Education Commission (KM202310009003)

More Information

Corresponding author: E-mail：zhangduona@buaa.edu.cn

摘要

摘要:
电磁环境日益复杂致使无线通信面临重大挑战，电磁信号调制识别作为认知无线电技术的关键环节意义深远。针对传统识别方法表征能力不足，深度学习方法适用性低、可解释性差的问题，结合2种方法的优点，提出领域知识内嵌的调制方式智能识别方法。所提方法将电磁信号高阶信息及频谱机制融入深度神经网络，提升分类性能及网络可解释程度。基于RML2018数据集，所提方法较ResNet调制识别正确率提升6.31%。
- 电磁特征 /
- 深度神经网络 /
- 调制方式识别 /
- 无线通信 /
- 知识内嵌
Abstract:
With the increasingly complex electromagnetic environment, wireless communication is facing severe challenges, making modulation recognition of electromagnetic signals, which becomes an important aspect of cognitive radio technology. Deep learning techniques have poor interpretability and little applicability, while traditional identification techniques have limited representation capabilities. In this paper, we propose an intelligent modulation recognition method that combines the advantages of both methods by embedding domain knowledge. In order to enhance classification performance and network interpretability, this technique integrates deep neural networks with high-order information and electromagnetic signal spectrum processes. Based on the RML2018 dataset, our method achieves a 6.31% improvement in modulation recognition accuracy compared to the ResNet method.
- electromagnetic characteristics /
- deep neural network /
- modulation recognition /
- wireless communications /
- knowledge-embedded

HTML全文

图 1 领域知识内嵌的调制方式智能识别架构

Figure 1. Intelligent identification architecture of modulation modes embedded in domain knowledge

下载: 全尺寸图片幻灯片

图 2 高阶信息统计量提取模块

Figure 2. High-order information statistics extraction module

下载: 全尺寸图片幻灯片

图 3 多谱通道注意力机制

Figure 3. Multispectral channel attention mechanism

下载: 全尺寸图片幻灯片

图 4 数据集信号生成流程

Figure 4. Dataset signal generation process

下载: 全尺寸图片幻灯片

图 5 数据集可视化

Figure 5. Dataset visualization

下载: 全尺寸图片幻灯片

图 6 模型准确率对比图

Figure 6. Model accuracy contrast chart

下载: 全尺寸图片幻灯片

图 7 ResNet模型混淆矩阵

Figure 7. ResNet model confusion matrix

下载: 全尺寸图片幻灯片

图 8 MSNet模型混淆矩阵

Figure 8. MSNet model confusion matrix

下载: 全尺寸图片幻灯片

图 9 本文模型混淆矩阵

Figure 9. The proposed model confusion matrix

下载: 全尺寸图片幻灯片

图 10 本文模型信噪比0 dB的混淆矩阵

Figure 10. The proposed model in SNR 0 dB confusion matrix

下载: 全尺寸图片幻灯片

图 11 本文模型信噪比14 dB的混淆矩阵

Figure 11. The proposed model in SNR 14 dB confusion matrix

下载: 全尺寸图片幻灯片

图 12 本文模型信噪比20 dB的混淆矩阵

Figure 12. The proposed model in SNR 20 dB confusion matrix

下载: 全尺寸图片幻灯片

图 13 本文模型信噪比30 dB的混淆矩阵

Figure 13. The proposed model in SNR 30 dB confusion matrix

下载: 全尺寸图片幻灯片

图 14 消融实验

Figure 14. Ablation experiments

下载: 全尺寸图片幻灯片

表 1 网络结构参数

Table 1. Network structure parameters

网络层	卷积核	输出维度
输入层		1024×2
残差层	32，3×2	512×32
高阶卷积层	1，1×1	512×32
多谱注意力层		512×32
残差层	32，3×1	256×32
残差层	32，3×1	128×32
高阶卷积层	1，1×1	128×32
多谱注意力层		128×32
残差层	32，3×1	64×32
残差层	32，3×1	32×32
高阶卷积层	1，1×1	32×32
多谱注意力层		32×32
残差层	32，3×1	16×32
全连接层		128
全连接层		24

下载: 导出CSV

表 2 数据集包含调制类型

Table 2. Dataset contains the modulation type

分类	调制方式
模拟调制	AM-SSB-WC/AM-SSBSC/AM-DSB-WC/AM-DSB-SC/FM
数字调制	OOK/4ASK/8ASK/BPSK/QPSK/8PSK/16PSK/32PSK/16APSK/32APSK/64APSK/128APSK/16QAM/32QAM/64QAM/128QAM/256QAM/ /GMSK/OQPSK

下载: 导出CSV

参考文献(26)

[1]	BKASSINY M, JAYAWEERA S K, LI Y, et al. Wideband spectrum sensing and non-parametric signal classification for autonomous self-learning cognitive radios[J]. IEEE Transactions on Wireless Communications, 2012, 11(7): 2596-2605.
[2]	WEI W, MENDEL J M. Maximum-likelihood classification for digital amplitude-phase modulations[J]. IEEE Transactions on Communications, 2000, 48(2): 189-193.
[3]	PANAGIOTOU P, ANASTASOPOULOS A, POLYDOROS A. Likelihood ratio tests for modulation classification[C]//Proceedings of the 21st Century Military Communications. Architectures and Technologies for Information Superiority. Piscataway: IEEE Press, 2002: 670-674.
[4]	HAMEED F, DOBRE O A, POPESCU D C. On the likelihood-based approach to modulation classification[J]. IEEE Transactions on Wireless Communications, 2009, 8(12): 5884-5892.
[5]	LI S T, QUAN D Y, WANG X F, et al. LPI radar signal modulation recognition with feature fusion based on time- frequency transforms[C]//Proceedings of the 13th International Conference on Wireless Communications and Signal Processing. Piscataway: IEEE Press, 2021: 1-6.
[6]	NANDI A K, AZZOUZ E E. Modulation recognition using artificial neural networks[J]. Signal Processing, 1997, 56(2): 165-175.
[7]	NANDI A K, AZZOUZ E E. Algorithms for automatic modulation recognition of communication signals[J]. IEEE Transactions on Communications, 1998, 46(4): 431-436.
[8]	SHI F N, JING X J, HE Y, et al. Classification model of wireless signals based on higher-order statistics[C]//Proceedings of the IEEE International Symposium on Broadband Multimedia System and Broadcasting. Piscataway: IEEE Press, 2020: 1-5.
[9]	李跃, 郭兴吉, 赵欣. 基于高阶累积量的调制方式识别研究[J]. 西南科技大学学报, 2018, 33(3): 64-68. LI Y, GUO X J, ZHAO X. Study on modulation recognition based on higher-order cumulants[J]. Journal of Southwest University of Science and Technology, 2018, 33(3): 64-68(in Chinese).
[10]	RUAN G X, ZHAO L, LIU Z. Recognition of continuous phase modulation signals based on synchrosqueezing wavelet transform[C]//Proceedings of the International Applied Computational Electromagnetics Society Symposium. Piscataway: IEEE Press, 2022: 1-2.
[11]	谭晓衡, 褚国星, 张雪静, 等. 基于高阶累积量和小波变换的调制识别算法[J]. 系统工程与电子技术, 2018, 40(1): 171-177. TAN X H, CHU G X, ZHANG X J, et al. Modulation recognition algorithm based on high-order cumulants and wavelet transform[J]. Systems Engineering and Electronics, 2018, 40(1): 171-177(in Chinese).
[12]	HAN G, LI J D, LU D H. Study of modulation recognition based on HOCs and SVM[C]//Proceedings of the IEEE 59th Vehicular Technology Conference. Piscataway: IEEE Press, 2004: 898-902.
[13]	TENG X Y, TIAN P W, YU H Y. Modulation classification based on spectral correlation and SVM[C]//Proceedings of the 4th International Conference on Wireless Communications, Networking and Mobile Computing. Piscataway: IEEE Press, 2008: 1-4.
[14]	MUSTAFA H, DOROSLOVACKI M. Digital modulation recognition using support vector machine classifier[C]//Proceedings of the Conference Record of the Thirty-Eighth Asilomar Conference on Signals, Systems and Computers. Piscataway: IEEE Press, 2005: 2238-2242.
[15]	JAGANNATH J, POLOSKY N, O’CONNOR D, et al. Artificial neural network based automatic modulation classification over a software defined radio testbed[C]//Proceedings of the IEEE International Conference on Communications. Piscataway: IEEE Press, 2018: 1-6.
[16]	TAYAKOUT H, DAYOUB I, GHANEM K, et al. Automatic modulation classification for D-STBC cooperative relaying networks[J]. IEEE Wireless Communications Letters, 2018, 7(5): 780-783.
[17]	WAHLA A H, CHEN L, WANG Y L, et al. Automatic wireless signal classification in multimedia Internet of Things: an adaptive boosting enabled approach[J]. IEEE Access, 2019, 7: 160334-160344.
[18]	MA K, ZHOU Y B, CHEN J Y. CNN-based automatic modulation recognition of wireless signal[C]//Proceedings of the IEEE 3rd International Conference on Information Systems and Computer Aided Education. Piscataway: IEEE Press, 2020: 654-659.
[19]	邵敏兰, 周鸿渐, 张浩然. 基于卷积神经网络的调制信号识别算法[J]. 实验室研究与探索, 2021, 40(8): 28-31. SHAO M L, ZHOU H J, ZHANG H R. Modulation signal recognition algorithm based on convolutional neural network[J]. Research and Exploration in Laboratory, 2021, 40(8): 28-31(in Chinese).
[20]	查雄, 彭华, 秦鑫, 等. 基于循环神经网络的卫星幅相信号调制识别与解调算法[J]. 电子学报, 2019, 47(11): 2443-2448. ZHA X, PENG H, QIN X, et al. Satellite amplitude-phase signals modulation identification and demodulation algorithm based on the cyclic neural network[J]. Acta Electronica Sinica, 2019, 47(11): 2443-2448(in Chinese).
[21]	PANCHAPAGESAN S, PARK D S, CHIU C C, et al. Efficient knowledge distillation for RNN-transducer models[C]//Proceedings of the IEEE International Conference on Acoustics, Speech and Signal Processing. Piscataway: IEEE Press, 2021: 5639-5643.
[22]	DING Z. Matrix outer-product decomposition method for blind multiple channel identification[J]. IEEE Transactions on Signal Processing, 1997, 45(12): 3053-3061.
[23]	O’SHEA T J, ROY T, CLANCY T C. Over-the-air deep learning based radio signal classification[J]. IEEE Journal of Selected Topics in Signal Processing, 2018, 12(1): 168-179.
[24]	HU J, SHEN L, SUN G. Squeeze-and-excitation networks[C]//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. Piscataway: IEEE Press, 2018: 7132-7141.
[25]	LIU K, LI F. Automatic modulation recognition based on a multiscale network with statistical features[J]. Physical Communication, 2023, 58: 102052.
[26]	SUN S Q, WANG Y Y. A novel deep learning automatic modulation classifier with fusion of multichannel information using GRU[J]. EURASIP Journal on Wireless Communications and Networking, 2023, 2023(1): 66.