zjwfufu / AMC-Net

Official code for "AMC-Net: An Effective Network for Automatic Modulation Classification".

AMC-Net

Code for "AMC-Net: An Effective Network for Automatic Modulation Classification".

Jiawei Zhang, Tiantian Wang, Zhixi Feng, and Shuyuan Yang

Xidian University

[Paper] | [中文文档] | [code] | [poster] | [video]

Preparation

Data

We conducted experiments on three datasets, namely RML2016.10a and RML2016.10b.

dataset	modulation formats	samples
RML2016.10a	8 digital formats: 8PSK, BPSK, CPFSK, GFSK, PAM4, 16QAM, 64QAM, QPSK; 3 analog formats: AM-DSB，AM-SSB，WBFM	220 thousand (2×128)
RML2016.10b	8 digital formats: 8PSK, BPSK, CPFSK, GFSK, PAM4, 16QAM, 64QAM, QPSK; 3 analog formats: AM-DSB，WBFM	1.2 million (2×128)

The datasets can be downloaded from the DeepSig. Please extract the downloaded compressed file directly into the ./data directory, and keep the file name unchanged. The final directory structure of ./data should is shown below:

data
├── RML2016.10a_dict.pkl
└── RML2016.10b.dat

Pretrained Model

We provide pre-trained models on two datasets, which can be downloaded from Google Drive or Baidu Netdisk. Please extract the downloaded compressed file directly into the ./checkpoint directory.

Environment Setup

• Python >= 3.6
• PyTorch >=1.7

This version of the code has been tested on Pytorch==1.8.1.

Training & Evaluation

The whole pipeline is adopted from our another work AWN. You can find the details of training and evaluation in there.

Visualize

We provide an additional mode to visualize the signal before ACM and after ACM, which can be called by the following command:

python main.py --mode visualize --dataset <DATASET>

Similar to Evaluation, the plotted figures are stored in ./result in the form of .svg.

Surprisingly, if we input a batch of random noise, and use ACM autoregressively:

Its behavior looks like some kind of implicit generative model. This property may help to achieve online augmentation.

Future Work

• Extend AMC-Net on RadioML2018.01a (long sequences).
• Investigate the capability of ACM.

License

This code is distributed under an MIT LICENSE. Note that our code depends on other libraries and datasets which each have their own respective licenses that must also be followed.

Citation

Please consider citing our paper if you find it helpful in your research:

@misc{zhang2023amcnet,
      title={AMC-Net: An Effective Network for Automatic Modulation Classification}, 
      author={Jiawei Zhang and Tiantian Wang and Zhixi Feng and Shuyuan Yang},
      year={2023},
      eprint={2304.00445},
      archivePrefix={arXiv},
      primaryClass={eess.SP}
}

Contact at: zjw AT stu DOT xidian DOT edu DOT cn