Machine Learning and Intelligent Communications. Second International Conference, MLICOM 2017, Weihai, China, August 5-6, 2017, Proceedings, Part I

Research Article

A Deep Learning Method Based on Convolutional Neural Network for Automatic Modulation Classification of Wireless Signals

Download
548 downloads
  • @INPROCEEDINGS{10.1007/978-3-319-73564-1_37,
        author={Yu Xu and Dezhi Li and Zhenyong Wang and Gongliang Liu and Haibo Lv},
        title={A Deep Learning Method Based on Convolutional Neural Network for Automatic Modulation Classification of Wireless Signals},
        proceedings={Machine Learning and Intelligent Communications. Second International Conference, MLICOM 2017, Weihai, China, August 5-6, 2017, Proceedings, Part I},
        proceedings_a={MLICOM},
        year={2018},
        month={2},
        keywords={Modulation classification Deep learning Convolutional neural network Wireless signal},
        doi={10.1007/978-3-319-73564-1_37}
    }
    
  • Yu Xu
    Dezhi Li
    Zhenyong Wang
    Gongliang Liu
    Haibo Lv
    Year: 2018
    A Deep Learning Method Based on Convolutional Neural Network for Automatic Modulation Classification of Wireless Signals
    MLICOM
    Springer
    DOI: 10.1007/978-3-319-73564-1_37
Yu Xu1,*, Dezhi Li1,*, Zhenyong Wang,*, Gongliang Liu1,*, Haibo Lv1,*
  • 1: Harbin Institute of Technology
*Contact email: xu_yu@hit.edu.cn, lidezhi@hit.edu.cn, ZYWang@hit.edu.cn, liugl@hit.edu.cn, elitelv@hit.edu.cn

Abstract

Automatic modulation classification (AMC) plays an important role in many fields to identify the modulation type of wireless signals. In this paper, we introduce deep learning to signal recognition. Based on architecture analysis of the convolutional neural network (CNN), we used real signal data generated by instruments as dataset, and proposed an improved CNN architecture to achieve compatible recognition accuracy of modulation classification. According to various conditions of signal noise ratio (SNR), we test the proposed CNN architecture with the real sampled signals. Experiments results show that the high-layer network is not necessary for modulation recognition with high SNR signals. The proposed CNN architecture has higher average classification accuracy than RESNET and is more compatible for modulation classification of signals with lower SNR.