by Jialang Xu (e-mail: jialang.xu.22@ucl.ac.uk), Chunbo Luo, Gerard Parr, Yang Luo.

Official implement of the paper, 'A Spatiotemporal Multi-Channel Learning Framework for Automatic Modulation Recognition'.

This repository contains MCLDNN implementation and datasets in the paper.

The code of this repository is integrated into the AMR Benchmark, which provides a unified implementation of several baseline deep learning models for automatic modulation recognition, thanks to the great contribution of Fuxin Zhang.

Automatic modulation recognition (AMR) plays a vital role in modern communication systems. We proposes a novel three-stream deep learning framework to extract the features from individual and combined in-phase/quadrature (I/Q) symbols of the modulated data. The proposed framework integrates one-dimensional (1D) convolutional, two-dimensional (2D) convolutional and long short-term memory (LSTM) layers to extract features more effectively from a time and space perspective. Experiments on the benchmark dataset show the proposed framework has efficient convergence speed and achieves improved recognition accuracy, especially for the signals modulated by higher dimensional schemes such as 16 quadrature amplitude modulation (16-QAM) and 64-QAM.

If this work is useful for your research, please consider citing:

@ARTICLE{9106397,
	author={J. {Xu} and C. {Luo} and G. {Parr} and Y. {Luo}},
	journal={IEEE Wireless Communications Letters}, 
	title={A Spatiotemporal Multi-Channel Learning Framework for Automatic Modulation Recognition}, 
	year={2020},
	volume={9},
	number={10},
	pages={1629-1632},
	doi={10.1109/LWC.2020.2999453}
	}

The recognition accuracy of the MCLDNN is shown in Fig.1.

Fig.1 Recognition accuracy comparison on the RadioML2016.10a dataset.

The available datasets can be downloaded from the table below:

• Python 3.6.10
• TensorFlow-gpu 1.14.0
• Keras-gpu 2.2.4

For the RadioML2016.10a dataset:

python train.py --datasetpath /path/to/RML2016.10a_dict.pkl --data 0

For the RadioML2016.10b dataset:

python train.py --datasetpath /path/to/RML2016.10b.dat --data 1

For the RadioML2016.10a dataset:

python test.py --datasetpath /path/to/RML2016.10a_dict.pkl --data 0

For the RadioML2016.10b dataset:

python test.py --datasetpath /path/to/RML2016.10b.dat --data 1

Weights for the RML2016.10a dataset. [GitHub download]

Note that our code is partly based on radioml. Thanks leena201818 for his great work!