{"id":26397455,"url":"https://github.com/thalesgroup/pythagore-mod-reco","last_synced_at":"2025-03-17T12:17:53.127Z","repository":{"id":43375893,"uuid":"370287625","full_name":"ThalesGroup/pythagore-mod-reco","owner":"ThalesGroup","description":"Package to train and run modulation recognition on raw I/Q radio samples, via deep-learning models","archived":false,"fork":false,"pushed_at":"2023-01-13T12:02:29.000Z","size":446,"stargazers_count":18,"open_issues_count":0,"forks_count":11,"subscribers_count":5,"default_branch":"master","last_synced_at":"2024-04-16T00:06:38.836Z","etag":null,"topics":["data-augmentation","deep-learning","deep-neural-network","machine-learning","modulation-classification","modulation-recognition","radio"],"latest_commit_sha":null,"homepage":"","language":"Jupyter Notebook","has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":"mit","status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/ThalesGroup.png","metadata":{"files":{"readme":"README.md","changelog":"CHANGELOG.md","contributing":"CONTRIBUTING.md","funding":null,"license":"LICENSE","code_of_conduct":null,"threat_model":null,"audit":null,"citation":null,"codeowners":null,"security":"SECURITY.md","support":null}},"created_at":"2021-05-24T08:53:23.000Z","updated_at":"2024-04-08T09:06:53.000Z","dependencies_parsed_at":"2023-02-09T15:30:51.213Z","dependency_job_id":null,"html_url":"https://github.com/ThalesGroup/pythagore-mod-reco","commit_stats":null,"previous_names":[],"tags_count":0,"template":false,"template_full_name":"ThalesGroup/template-project","repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/ThalesGroup%2Fpythagore-mod-reco","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/ThalesGroup%2Fpythagore-mod-reco/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/ThalesGroup%2Fpythagore-mod-reco/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/ThalesGroup%2Fpythagore-mod-reco/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/ThalesGroup","download_url":"https://codeload.github.com/ThalesGroup/pythagore-mod-reco/tar.gz/refs/heads/master","host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":244031128,"owners_count":20386534,"icon_url":"https://github.com/github.png","version":null,"created_at":"2022-05-30T11:31:42.601Z","updated_at":"2022-07-04T15:15:14.044Z","host_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub","repositories_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories","repository_names_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repository_names","owners_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners"}},"keywords":["data-augmentation","deep-learning","deep-neural-network","machine-learning","modulation-classification","modulation-recognition","radio"],"created_at":"2025-03-17T12:17:52.153Z","updated_at":"2025-03-17T12:17:53.112Z","avatar_url":"https://github.com/ThalesGroup.png","language":"Jupyter Notebook","funding_links":[],"categories":[],"sub_categories":[],"readme":"# pythagore-mod-reco\n\nModulation recognition AI algorithms benchmark.\\\nThis project contains a Jupyter Notebook for the interactive benchmark, deep learning networks and a few utility functions gathered into a package.\n\n## Package to run modulation recognition on raw I/Q radio samples\n\nThe acompagning paper is: \"A light neural network for modulation detection under impairments, T. Courtat, H. du Mas des Bourboux, 2021\"\npresented at the \"2021 International Symposium on Networks, Computers and Communications (ISNCC'21)\"\n(\u003chttp://www.isncc-conf.org/\u003e).\n\nThe dataset, and a notebook to reproduce the results can be found on Kaggle: \u003chttps://www.kaggle.com/hdumasde/pythagoremodreco\u003e\n\nExample: evolution of the error rate with the number of epochs on the AugMod dataset |\n:-----------------------------------------------------------------------------------:|\n![example-training](/sources/network_comparison_AugMod-trained1024.png)              |\n\n## Setup\n\nThe package requires Python \u003e=3.7.\n\nIt has been tested on Nvidia GPU with cuda 10.2.\n\n### Install package and dependencies\n\nThe package installation is as simple as\n\n```bash\ncd \u003cproject_folder\u003e\npip3 install -U .\n```\n\nIf you only need to install dependencies you can go with\n\n```bash\npython -m pip install -r requirements.txt\n```\n\n### Get data\n\nThe trainning and testing of algorithms can be performed on several datasets:\n\n- The pythagore-mod-reco AugMod dataset as presented in our article:\n\n  - can be found on Kaggle: \u003chttps://www.kaggle.com/hdumasde/pythagoremodreco\u003e\n\n  - can be downloaded with wget\n\n```bash\nwget https://augmod.blob.core.windows.net/augmod/augmod.zip\nunzip augmod.zip\n```\n\n- RML 2016 datasets from DeepSig:\n\n```bash\nwget https://opendata.deepsig.io/datasets/2016.04/2016.04C.multisnr.tar.bz2?__hstc=233546881.9c91e0549f9b6bfce6708a49c211c1c9.1614872457734.1614872457734.1614872457734.1\u0026__hssc=233546881.1.1614872457735\u0026__hsfp=1843090487\nwget https://opendata.deepsig.io/datasets/2016.10/RML2016.10b.tar.bz2?__hstc=233546881.9c91e0549f9b6bfce6708a49c211c1c9.1614872457734.1614872457734.1614872457734.1\u0026__hssc=233546881.1.1614872457735\u0026__hsfp=1843090487\nwget https://opendata.deepsig.io/datasets/2016.10/RML2016.10a.tar.bz2?__hstc=233546881.9c91e0549f9b6bfce6708a49c211c1c9.1614872457734.1614872457734.1614872457734.1\u0026__hssc=233546881.1.1614872457735\u0026__hsfp=1843090487\n```\n\n- RML 2018 dataset from DeepSig:\n\nTo get RadioML2018.01A  you should connect to \u003chttps://www.deepsig.ai/datasets\u003e\n\nPlease update the following paths: `data_path` and `log_path` in `jupyter/train-test-modulationreco.ipynb`, to where data are located and to where you want to store log files.\n\n## Run\n\n- Train and test the different networks on each datasets running `jupyter/train-test-modulationreco.ipynb`.\n\n## Source\n\n- The implementation of RML-ConvNet is found at \u003chttps://github.com/radioML/examples/blob/master/modulation_recognition/RML2016.10a_VTCNN2_example.ipynb\u003e\n- The implementation of RML-CNN/VGG is found at \u003chttps://github.com/leena201818/radioml/blob/master/rmlmodels/VGGLikeModel.py\u003e\n- The implementation of RML-ResNet is found at \u003chttps://github.com/liuzhejun/ResNet-for-Radio-Recognition/blob/master/ResNet_Model.ipynb\u003e\n\n## Citing\n\n- Please cite the following paper\nif you are using the AugMod dataset or Mod-LCNN or Mod-LRCNN networks\n\n```bibtex\n@INPROCEEDINGS{9615851,\n       author = {{Courtat}, Thomas and {du Mas des Bourboux}, H{\\'e}lion},\n    booktitle = {2021 International Symposium on Networks, Computers and Communications (ISNCC)},\n        title = {A light neural network for modulation detection under impairments},\n         year = {2021},\n       volume = {},\n       number = {},\n        pages = {1-7},\n          doi = {10.1109/ISNCC52172.2021.9615851},\narchivePrefix = {arXiv},\n       eprint = {2003.12260},\n primaryClass = {cs.LG},\n}\n```\n\n- Please visit \u003chttps://www.deepsig.ai/datasets\u003e to see how to cite RadioML datasets and networks\n\n## License\n\n- This repository is licensed under the terms of the MIT License (see the file LICENSE).\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fthalesgroup%2Fpythagore-mod-reco","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fthalesgroup%2Fpythagore-mod-reco","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fthalesgroup%2Fpythagore-mod-reco/lists"}