singingkettle/changshuoradiorecognition

AI Framework of Radio Recognition

Introduction

English | 简体中文

CSRR (ChangShuoRadioRecognition) is an open source signal processing toolbox based on PyTorch. The framework of this project is based on the mmdetection and mmcv.

Major features

• Support of multiple frameworks out of box

  The toolbox directly supports popular and contemporary AMC algorithms, e.g. CLDNN, TanhNet, etc.

Changelog

v2.0.1 was released in 8/8/2022.

v1.0.1 was released in 5/9/2020.

Benchmark and model zoo

Supported Automatic Modulation Classification methods:

• CGDNet. CGDNet: Efficient Hybrid Deep Learning Model for Robust Automatic Modulation Recognition
• CLDNNW. Deep architectures for modulation recognition
• CLDNNL. Deep Neural Network Architectures for Modulation Classification
• CNN1DPF. Automatic Modulation Classification Using Parallel Fusion of Convolutional Neural Networks
• CNN2. Convolutional Radio Modulation Recognition Networks
• CNN4. Robust and Fast Automatic Modulation Classification with CNN under Multipath Fading Channels
• DAE. Real-Time Radio Technology and Modulation Classification via an LSTM Auto-Encoder
• DensCNN. Deep Neural Network Architectures for Modulation Classification
• DSCLDNN. Automatic Modulation Classification Using CNN-LSTM Based Dual-Stream Structure
• FastMLDNN. A Fast Multi-Loss Learning Deep Neural Network for Automatic Modulation Classification
• GRU2. Automatic modulation classification using recurrent neural networks
• HCGDNN. A Hierarchical Classification Head based Convolutional Gated Deep Neural Network for Automatic Modulation Classification
• LSTM2. Deep Learning Models for Wireless Signal Classification With Distributed Low-Cost Spectrum Sensors
• IC-AMCNet. CNN-Based Automatic Modulation Classification for Beyond 5G Communications
• MCformer. MCformer: A Transformer Based Deep Neural Network for Automatic Modulation Classification
• MCLDNN. A Spatiotemporal Multi-Channel Learning Framework for Automatic Modulation Recognition
• MCNET. MCNet: An Efficient CNN Architecture for Robust Automatic Modulation Classification
• MLDNN. Multitask-Learning-Based Deep Neural Network for Automatic Modulation Classification
• FMLDNN. A Fast Multi-Loss Learning Deep Neural Network for Automatic Modulation Classification
• PET-CGDNN. An Efficient Deep Learning Model for Automatic Modulation Recognition Based on Parameter Estimation and Transformation
• TRN. Signal Modulation Classification Based on the Transformer Network

Installation

Please refer to install.md for installation and dataset preparation.

Getting Started

Please see getting_started.md for the basic usage of CSRR.

Benchmark Results

Please see summary.md for the benchmark results.

Version Control

For version control, we use the git. Please refer This tutorial

Citation

If you use this toolbox or benchmark in your research, please cite this project.

@article{csrr,
  title   = {{CSRR}: Open ChangShuo RadioRecognition Toolbox and Benchmark},
  author  = {Shuo Chang},
  journal= {coming soon},
  year={2020}
}

@article{chang2021multi,
  title={Multi-task learning based deep neural network for automatic modulation classification},
  author={Chang, Shuo and Huang, Sai and Zhang, Ruiyun and Feng, Zhiyong and Liu, Liang},
  journal={IEEE Internet of Things Journal},
  year={2021},
  publisher={IEEE}
}
@ARTICLE{9764618,
  author={Chang, Shuo and Zhang, Ruiyun and Ji, Kejia and Huang, Sai and Feng, Zhiyong},
  journal={IEEE Transactions on Wireless Communications}, 
  title={A Hierarchical Classification Head based Convolutional Gated Deep Neural Network for Automatic Modulation Classification}, 
  year={2022},
  volume={},
  number={},
  pages={1-1},
  doi={10.1109/TWC.2022.3168884}}

Acknowledgement

CSRR is an open source project that is contributed by ShuoChang. We appreciate all the contributors who implement their methods or add new features, as well as users who give valuable feedbacks. We wish that the toolbox and benchmark could serve the growing research community by providing a flexible toolkit to reimplement existing methods and develop their own new signal processing algorithms.

Contact

This repo is currently maintained by Shuo Chang (@Singingkettle),

changshuoradiorecognition's Issues

About mldnn and cache

Hello! I attempted to use mldnn on the Deepsig2018 dataset, but I removed parts with SNR greater than 20dB from the JSON file and then ran cache_ amc.py generated a series of .pkl files. Afterwards, I tried to train mldnn, but the loss was displayed as a straight line and there was no change in acc at all; As a comparison, I trained ResCNN in the same way, but it converged correctly. I noticed the use of cache in mldnn, is this the reason for this phenomenon? Did I do something wrong? thank you!

关于comvert_deepsig.py的问题

您好，当我用tools/convert_datasets/comvert_deepsig.py对RML2016.10A进行处理的时候，我发现程序会报错。
错误发生在deepsig.py的第88行：
item = data[(mod, snr)][sub_item_index,:,:]
其中mod变量在上面的77行被 mod = mod.decode('UTF-8')解码过，因此在88行中，data无法被str变量‘mod’访问。
将88行修改为：data[(mod.encode(), snr)][sub_item_index,:,:]可以解决这个问题

How can I draw result figure?

Hello!
I want to draw:

1. The relationship curve of SNR-Accuracy;
2. The confusion matrix of AMC recognition under a certain SNR

What commands should I use separately? Or do I need to implement it myself? thank you!

the class center distance expansion loss of FastMLDNN

I have read your paper about FastMLDNN. This is really excellent work. When I looked up your code, it seemed that I couldn't find the code of the class center distance expansion loss. Perhaps I missed it?

关于HCGDNN

我在summary.md中没有看到关于其的result，请问能否公开具体的snr-acc数值，从而方便比较呢？谢谢！