Proceedings of the 13th EAI International Conference on Mobile Multimedia Communications, Mobimedia 2020, 27-28 August 2020, Cyberspace

Research Article

Signal Modulation Recognition Method based on Time-frequency Image

Download668 downloads
  • @INPROCEEDINGS{10.4108/eai.27-8-2020.2294229,
        author={Yuqian  Li and Cheng  Chang and Chengzhuo  Shi and Sen  Wang},
        title={Signal Modulation Recognition Method based on Time-frequency Image},
        proceedings={Proceedings of the 13th EAI International Conference on Mobile Multimedia Communications, Mobimedia 2020, 27-28 August 2020, Cyberspace},
        publisher={EAI},
        proceedings_a={MOBIMEDIA},
        year={2020},
        month={11},
        keywords={modulation recognition time-frequency cnn image processing},
        doi={10.4108/eai.27-8-2020.2294229}
    }
    
  • Yuqian Li
    Cheng Chang
    Chengzhuo Shi
    Sen Wang
    Year: 2020
    Signal Modulation Recognition Method based on Time-frequency Image
    MOBIMEDIA
    EAI
    DOI: 10.4108/eai.27-8-2020.2294229
Yuqian Li1, Cheng Chang2, Chengzhuo Shi3, Sen Wang1,*
  • 1: Harbin Engineering University
  • 2: China Academy of Launch Vehicle Technology
  • 3: Nanjing Electronic Equipment Institute
*Contact email: wangsen@hrbeu.edu.cn

Abstract

Signal modulation classification is an important technology for signal processing, in order to get higher recognition accuracy at low signal-to-noise ratios, this paper proposed a modulation recognition algorithm which combines CNN and time-frequency analysis methods. The method’s steps are as following, first, make the SPWVD transforms to the digital signals to obtain corresponding time-frequency images. Then, use image processing method to enhance the obtained images, such as image gray processing, perform the grayscale equalization and binarization the images. After that, the pictures are inputted into the CNN network for classification and recognition. The experimental results show that the recognition rate can achieve 90.44% at 0dB.