Skip to content

Code for "Complex-Valued Depthwise Separable Convolutional Neural Network for Automatic Modulation Classification"

Notifications You must be signed in to change notification settings

Patrick-Nick/CDSCNN

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

8 Commits
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

CDSCNN

Complex-Valued Depthwise Separable Convolutional Neural Network for Automatic Modulation Classification

Code for "Complex-Valued Depthwise Separable Convolutional Neural Network for Automatic Modulation Classification". [Paper]

Chenghong Xiao, Shuyuan Yang, Senior Member, IEEE and ZhixiFeng, Member, IEEE

Abstract—Automatic modulation classification (AMC) is a critical task in industrial cognitive communication systems. Existing state-of-the-art methods, typified by real-valued convolutional neural networks, have introduced innovative solutions for AMC. However, such models viewed the two constituent components of complex-valued modulated signals as discrete real-valued inputs, causing structural phase damage to original signals and reduced interpretability of the model. In this article, a novel end-to-end AMC model called a complex-valued depthwise separable convolutional neural network (CDSCNN) is proposed, which adopts complex-valued operation units to enable automatic complex-valued feature learning specifically tailored for AMC. Considering the limited hardware resources available in industrial scenarios, complex-valued depthwise separable convolution (CDSC) is designed to strike a balance between classification accuracy and model complexity. With an overall accuracy (OA) of 62.63% on the RadioML2016.10a dataset, CDSCNN outperforms its counterparts by 1%–11%. After finetuning on the RadioML2016.10b dataset, the OA reaches 63.15%, demonstrating the robust recognition and generalization capability of CDSCNN. Moreover, the CDSCNN exhibits lower model complexity compared to other methods.

Datasets

We conducted experiments on two datasets, namely RadioML2016.10a, and RadioML2016.10b.

dataset modulation formats samples
RadioML2016.10a 8 digital formats: 8PSK, BPSK, CPFSK, GFSK, PAM4, 16QAM, 64QAM, QPSK; 3 analog formats: AM-DSB, AM-SSB, WBFM 220 thousand (2×128)
RadioML2016.10b 8 digital formats: 8PSK, BPSK, CPFSK, GFSK, PAM4, 16QAM, 64QAM, QPSK; 2 analog formats: AM-DSB, WBFM 1.2 million (2×128)

Requirements

  • python == 3.10.4
  • pytorch == 1.12.0
  • scikit-learn == 1.3.0
  • numpy == 1.21.5

Citation

Please consider citing our paper if you find it helpful in your research. Please do not hesitate to contact us (Email: ch_xiao@stu.xidian.edu.cn) if there are any problems.

@ARTICLE{10198896,
	author={Xiao, Chenghong and Yang, Shuyuan and Feng, Zhixi},
	journal={IEEE Transactions on Instrumentation and Measurement}, 
	title={Complex-Valued Depthwise Separable Convolutional Neural Network for Automatic Modulation Classification}, 
	year={2023},
	volume={72},
	pages={1-10},
	doi={10.1109/TIM.2023.3298657}
}