高级搜索

留言板

尊敬的读者、作者、审稿人, 关于本刊的投稿、审稿、编辑和出版的任何问题, 您可以本页添加留言。我们将尽快给您答复。谢谢您的支持!

姓名
邮箱
手机号码
标题
留言内容
验证码

基于单通道多尺度图神经网络的自动调制识别

国强 聂孟允 戚连刚 Kaliuzhnyi Mykola

国强, 聂孟允, 戚连刚, Kaliuzhnyi Mykola. 基于单通道多尺度图神经网络的自动调制识别[J]. 电子与信息学报, 2023, 45(5): 1575-1584. doi: 10.11999/JEIT220840
引用本文: 国强, 聂孟允, 戚连刚, Kaliuzhnyi Mykola. 基于单通道多尺度图神经网络的自动调制识别[J]. 电子与信息学报, 2023, 45(5): 1575-1584. doi: 10.11999/JEIT220840
GUO Qiang, NIE Mengyun, QI Liangang, Kaliuzhnyi Mykola. Automatic Modulation Recognition Based on Single-channel Multi-scale Graph Neural Network[J]. Journal of Electronics & Information Technology, 2023, 45(5): 1575-1584. doi: 10.11999/JEIT220840
Citation: GUO Qiang, NIE Mengyun, QI Liangang, Kaliuzhnyi Mykola. Automatic Modulation Recognition Based on Single-channel Multi-scale Graph Neural Network[J]. Journal of Electronics & Information Technology, 2023, 45(5): 1575-1584. doi: 10.11999/JEIT220840

基于单通道多尺度图神经网络的自动调制识别

doi: 10.11999/JEIT220840
基金项目: 国家重点研发计划(2018YFE0206500),国家自然科学基金(62071140),中央高校基本科研业务费专项资金(3072022QBZ0801)
详细信息
    作者简介:

    国强:男,教授,研究方向为电子对抗、智能信号处理与识别

    聂孟允:男,硕士生,研究方向为基于深度学习的信号识别

    戚连刚:男,讲师,研究方向为自适应干扰抑制、卫星导航信号处理

    Kaliuzhnyi Mykola:男,教授,研究方向为电子对抗、电磁目标识别

    通讯作者:

    戚连刚 qiliangang@hrbeu.edu.cn

  • 中图分类号: TN911

Automatic Modulation Recognition Based on Single-channel Multi-scale Graph Neural Network

Funds: The National Key R & D Plan(2018YFE0206500), The National Natural Science Foundation of China (62071140), The Fundamental Research Funds for Central Universities (3072022QBZ0801)
  • 摘要: 针对自适应可见性图(AVG)算法复杂度过高且精度提升不明显的缺点,该文提出一种基于单通道多尺度图神经网络(SMGNN)的自动调制识别(AMR)框架,并对框架各个部分进行了可解释性研究。首先利用多层感知机和1维卷积自适应地实现了单通道信号序列和图之间的映射,有效降低了AVG算法的复杂度;其次,设计了一种多尺度图神经网络,将不同分辨率的特征进行融合,提升了模型识别准确率。实验表明,该文提出的SMGNN算法相比于AVG算法节省了近1/2的参数量,且识别精度得到了较大的提升。
  • 图  1  基于单通道多尺度图神经网络的自动调制识别

    图  2  通道融合网络及激活函数

    图  3  自适应图表示网络框架

    图  4  多尺度图神经网络

    图  5  矢量调制图

    图  6  VG算法示意图

    图  7  基于T-SNE的可视化分析

    图  8  不同超参数m下的识别率

    图  9  不同算法的识别精度

    图  10  不同算法的混淆矩阵分析

    图  11  消融实验的混淆矩阵分析

    图  12  单通道可见性图

    表  1  数据集RadioML2016.10b的参数设置

    参数定义值
    采样速率(kHz)200
    信噪比(dB)–20~18,间隔为2
    数据格式IQ数据格式:2×128
    信号数量1 200 000
    下载: 导出CSV

    表  2  不同算法对数据集RML2016.10b的识别性能

    算法识别精度(%)参数量平均推理时间(s)
    CNN2D55.4127069540.0192
    RESNET63.95 1919460.0053
    LSTM63.721976420.0064
    MCLDNN64.454037220.0160
    AVGNET64.6123119840.0310
    SMGNN65.3612467200.0176
    下载: 导出CSV

    表  3  消融实验识别性能

    算法识别精度(%)参数量平均推理时间(s)
    SMGNNo(单通道单尺度)64.8211563520.0165
    SMGNNs(48维节点特征)64.817129760.0165
    SMGNNs(32维节点特征)64.673277120.0154
    SMGNNs(16维节点特征)64.48909280.0149
    下载: 导出CSV
  • [1] LIANG Yingchang, TAN Junjie, and NIYATO D. Overview on intelligent wireless communication technology[J]. Journal on Communications, 2020, 41(7): 1–17. doi: 10.11959/j.issn.1000-436x.2020145
    [2] XU J L, SU Wei, and ZHOU Mengchu. Likelihood-Ratio approaches to automatic modulation classification[J]. IEEE Transactions on Systems, Man, and Cybernetics, Part C, 2011, 41(4): 455–469. doi: 10.1109/TSMCC.2010.2076347
    [3] XU J L, SU Wei, and ZHOU Mengchu. Software-Defined radio equipped with rapid modulation recognition[J]. IEEE Transactions on Vehicular Technology, 2010, 59(4): 1659–1667. doi: 10.1109/TVT.2010.2041805
    [4] O’SHEA T J, CORGAN J, and CLANCY T C. Convolutional radio modulation recognition networks[C]. 17th International Conference on Engineering Applications of Neural Networks, Aberdeen, UK, 2016: 213–226.
    [5] WEST N E and O'SHEA T. Deep architectures for modulation recognition[C]. 2017 IEEE International Symposium on Dynamic Spectrum Access Networks (DySPAN), Baltimore, USA, 2017: 1–6.
    [6] 郭业才, 姚文强. 基于信噪比分类网络的调制信号分类识别算法[J]. 电子与信息学报, 2022, 44(10): 3507–3515. doi: 10.11999/JEIT210825

    GUO Yecai and YAO Wenqiang. Modulation signal classification and recognition algorithm based on signal to noise ratio classification network[J]. Journal of Electronics &Information Technology, 2022, 44(10): 3507–3515. doi: 10.11999/JEIT210825
    [7] 颜康, 金炜东, 黄颖坤, 等. 基于元学习的畸变雷达电磁信号识别[J]. 电子与信息学报, 2022, 44(4): 1351–1357. doi: 10.11999/JEIT210190

    YAN Kang, JIN Weidong, HUANG Yingkun, et al. Distorted radar electromagnetic signal recognition based on meta-learning[J]. Journal of Electronics &Information Technology, 2022, 44(4): 1351–1357. doi: 10.11999/JEIT210190
    [8] XUE Fuzhao, SUN Aixin, ZHANG Hao, et al. GDPNet: Refining latent multi-view graph for relation extraction[J]. Proceedings of the AAAI Conference on Artificial Intelligence, 2021, 35(16): 14194–14202. doi: 10.1609/aaai.v35i16.17670
    [9] NAN Guoshun, LUO Guoqing, LENG Sicong, et al. Speaker-oriented latent structures for dialogue-based relation extraction[J]. arXiv preprint arXiv: 2109.05182, 2021.
    [10] LACASA L, LUQUE B, BALLESTEROS F, et al. From time series to complex networks: The visibility graph[J]. Proceedings of the National Academy of Sciences of the United States of America, 2008, 105(13): 4972–4975. doi: 10.1073/pnas.0709247105
    [11] LUQUE B, LACASA L, BALLESTEROS F, et al. Horizontal visibility graphs: Exact results for random time series[J]. Physical Review E, 2009, 80(4 Pt 2): 046103.
    [12] QI Xuan, ZHOU Jinchao, QIU Kunfeng, et al. CLPVG: Circular limited penetrable visibility graph as a new network model for time series[J]. Chaos:An Interdisciplinary Journal of Nonlinear Science, 2022, 32(1): 013130. doi: 10.1063/5.0048243
    [13] WAN Tao, JIANG Kaili, TANG Yanli, et al. Automatic LPI radar signal sensing method using visibility graphs[J]. IEEE Access, 2020, 8: 159650–159660. doi: 10.1109/ACCESS.2020.3020336
    [14] LIU Yabo, LIU Yi, and YANG Cheng. Modulation recognition with graph convolutional network[J]. IEEE Wireless Communications Letters, 2020, 9(5): 624–627. doi: 10.1109/LWC.2019.2963828
    [15] QI Xuan, ZHOU Jinchao, QIU Kunfeng, et al. AvgNet: Adaptive visibility graph neural network and its application in modulation classification[J]. IEEE Transactions on Network Science and Engineering, 2022, 9(3): 1516–1526. doi: 10.1109/TNSE.2022.3146836
    [16] YING R, YOU Jiaxuan, MORRIS C, et al. Hierarchical graph representation learning with differentiable pooling[C]. Proceedings of the 32nd International Conference on Neural Information Processing Systems, Montréal, Canada, 2018: 4805–4815.
    [17] VAN DER MAATEN L and HINTON G. Visualizing data using t-SNE[J]. Journal of Machine Learning Research, 2008, 9(2605): 2579–2605.
    [18] 樊昌信, 曹丽娜. 通信原理[M]. 7版. 北京: 国防工业出版社, 2012: 176–225.

    FAN Changxin and CAO Lina. Principles of Communication[M]. 7th ed. Beijing: National Defense Industry Press, 2012: 176–225.
    [19] LIN T Y, DOLLAR P, GIRSHICK R, et al. Feature pyramid networks for object detection[C]. The 2017 IEEE Conference on Computer Vision and Pattern Recognition, Honolulu, USA, 2017: 936–944.
    [20] O’SHEA T J, ROY T, and 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. doi: 10.1109/JSTSP.2018.2797022
    [21] RAJENDRAN S, MEERT W, GIUSTINIANO D, et al. Deep learning models for wireless signal classification with distributed low-cost spectrum sensors[J]. IEEE Transactions on Cognitive Communications and Networking, 2018, 4(3): 433–445. doi: 10.1109/TCCN.2018.2835460
    [22] XU Jialang, LUO Chunbo, PARR G, et al. A spatiotemporal multi-channel learning framework for automatic modulation recognition[J]. IEEE Wireless Communications Letters, 2020, 9(10): 1629–1632. doi: 10.1109/LWC.2020.2999453
    [23] BASTIAN M, HEYMANN S, and JACOMY M. Gephi: An open source software for exploring and manipulating networks[J]. Proceedings of the International AAAI Conference on Web and Social Media, 2009, 3(1): 361–362. doi: 10.1609/icwsm.v3i1.13937
  • 加载中
图(12) / 表(3)
计量
  • 文章访问数:  592
  • HTML全文浏览量:  245
  • PDF下载量:  173
  • 被引次数: 0
出版历程
  • 收稿日期:  2022-06-24
  • 修回日期:  2022-11-08
  • 网络出版日期:  2022-11-10
  • 刊出日期:  2023-05-10

目录

    /

    返回文章
    返回