Study On Optical Communications, Volume. 49, Issue 5, 16(2023)

Joint PAPR Reduction Technology Enabled by Neural Network-based Coding and Companding

Jun-yuan Nie1... Da-wei Zhang1,* and Jing Zhang2 |Show fewer author(s)
Author Affiliations
  • 1School of Electronic Information and Communication, Huazhong University of Science and Technology, Wuhan 430074, China
  • 2Key Lab.of Optical Fiber Sensing and Communications, University of Electronic Science and Technology, Chengdu 611731, China
  • show less
    Figures & Tables(11)
    OFDM schematic diagram
    FT-PRNet schematic diagram
    BER vs SNR at spreading ratio(0,10%,25%)
    BER at the subcarrier number of(16,64,256,1 024)
    BER performance comparison with different schemes
    Test block diagram
    The relationship between PAPR and BER in measurement
    • Table 1. Model architecture parameters

      View table
      View in Article

      Table 1. Model architecture parameters

      模块维度(输入,输出)
      频域编码全连接(2×(1-anoc, 2×noc)
      时域压扩IFFT+滑窗-
      全连接+Relu(2×3,32)
      全连接+Relu(32,32)
      全连接(32,2)
      噪声2
      时域解压滑窗-
      全连接+Relu(2×3,64)
      全连接+Relu(64,64)
      全连接(64,2)
      FFT-
      频域解码全连接(2×noc, 2×(1-anoc)
    • Table 2. Basic parameters

      View table
      View in Article

      Table 2. Basic parameters

      参数数值设置
      比特数/符号k=4
      信噪比/dB8~24
      子载波数16,64,256,1 024
      权重λ0.05
      扩频比例0,10%,25%
      训练集/测试集60 000/1 024 OFDM符号
      优化器/学习率自适应矩估计(Adam)/5e-4
      批大小64
    • Table 3. Performance comparison with different schemes

      View table
      View in Article

      Table 3. Performance comparison with different schemes

      方案SNR=8~24 dB对应的CCDFCCDF=0.100,0.010,0.001 PAPR分别减少值
      Mu-law
      TNN
      PRNet
      FT-PRNet
    • Table 4. Parameters of main device

      View table
      View in Article

      Table 4. Parameters of main device

      发射端参数接收端参数
      上位机信号带宽下位机信号带宽
      800 Mbit/s800 Mbit/s
      AWG带宽=30 GHzDSO带宽=30 GHz
      采样率64 GS/s采样率80 GS/s
      天线发工作频段天线收工作频段
      26.3~40.0 GHz26.3~40.0 GHz
    Tools

    Get Citation

    Copy Citation Text

    Jun-yuan Nie, Da-wei Zhang, Jing Zhang. Joint PAPR Reduction Technology Enabled by Neural Network-based Coding and Companding[J]. Study On Optical Communications, 2023, 49(5): 16

    Download Citation

    EndNote(RIS)BibTexPlain Text
    Save article for my favorites
    Paper Information

    Category: Research Articles

    Received: May. 1, 2022

    Accepted: --

    Published Online: Nov. 22, 2023

    The Author Email: Zhang Da-wei (dwzhang@hust.edu.cn)

    DOI:10.13756/j.gtxyj.2023.05.003

    Topics