High Power Laser and Particle Beams, Volume. 36, Issue 4, 043016(2024)

Radar radiation source recognition method based on compressed residual network

Enze Guo1, Zhengtang Liu1, Bo Cui1, Guobin Liu1, Hangyu Shi2, and Xu Jiang1、*
Author Affiliations
  • 1Unit 63893 of the PLA, Luoyang 471003, China
  • 2Unit 63896 of the PLA, Luoyang 471003, China
  • show less
    Figures & Tables(11)
    CWD time-frequency distribution of 14 radar emitter signals
    Signal preprocessing flow chart
    Basic structure of residual unit
    Expansion convolution and receptive field
    CRN model
    Comparison of recognition rates of different design methods
    Recognition rate of polyphase coded signals by different methods
    Confusion matrix for different methods to identify polyphase coding
    Overall recognition rate of each model
    • Table 1. Signal parameter setting

      View table
      View in Article

      Table 1. Signal parameter setting

      modulation typeparameterranges
      CWcarrier frequency $ {f_0} $[1/10,1/4]fs
      LFM、NLFMcarrier frequency $ {f_0} $[1/10,1/4]fs
      bandwidth B[1/10,1/4]fs
      BPSKcarrier frequency $ {f_0} $[1/10,1/4]fs
      barker code length7
      QPSKcarrier frequency $ {f_0} $[1/10,1/4]fs
      phase coding sequence length6
      FMCWcarrier frequency $ {f_0} $[1/20,1/5]fs
      bandwidth B[1/20,1/5]fs
      cycle T1$ {\text{μ s}} $
      FRANK、P1, P2, P3, P4carrier frequency $ {f_0} $[1/10,1/4]fs
      step frequency N{6,8}
      BFSKcarrier frequency $ {f_0} $[1/20,1/4]fs
      barker code length{11,13}
      QFSKcarrier frequency $ {f_0} $[1/10,1/4]fs
      frequency coding sequence length{6,7}
      COSTAScarrier frequency $ {f_0} $[1/10,1/4]fs
      frequency coding sequence length{6,7,8}
    • Table 2. Multi-adds calculation amount required for each convolution operation of different models

      View table
      View in Article

      Table 2. Multi-adds calculation amount required for each convolution operation of different models

      model structureFLOPS
      model 2 of this articleRef. [22] model
      conv1112×112×1×32×7×719668992112×112×1×32×7×719668992
      conv2_x56×56×32×32×3×3×25780275256×56×32×32×3×3×257802752
      conv3_x28×28×32×64×3×3+28×28×64×64×3×34335206428×28×32×64×3×3+28×28×64×64×3×343352064
      conv4_x14×14×64×128×3×3+14×14×128×128×3×34335206428×28×64×128×3×3+28×28×128×128×3×3173408256
      conv5_x7×7×128×256×3×3+7×7×256×256×3×34335206428×28×128×256×3×3+28×28×256×256×3×3693633024
      conv6_x3×3×256×256×3×3×21061683228×28×256×256×3×3×2924844032
      fc256×143584256×143584
      sum218, 148, 3521, 912, 712, 704
    Tools

    Get Citation

    Copy Citation Text

    Enze Guo, Zhengtang Liu, Bo Cui, Guobin Liu, Hangyu Shi, Xu Jiang. Radar radiation source recognition method based on compressed residual network[J]. High Power Laser and Particle Beams, 2024, 36(4): 043016

    Download Citation

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

    Category:

    Received: May. 6, 2023

    Accepted: Oct. 20, 2023

    Published Online: Apr. 22, 2024

    The Author Email: Xu Jiang (13525965959@139.com)

    DOI:10.11884/HPLPB202436.230119

    Topics