This repository curates papers and code for applying deep learning to wireless communication problems, targeting researchers and engineers in the field. It aims to bridge the gap caused by the lack of open-source code in many wireless communication publications, fostering reproducible research and accelerating learning.

How It Works

The project systematically collects and categorizes papers that combine deep learning with wireless communications, specifically focusing on those that provide publicly available source code. It leverages a combination of web scraping and GitHub Actions for daily updates on new research in relevant arXiv categories (e.g., eess.SP, cs.IT). The organization into topics like physical layer optimization, resource management, and emerging systems provides a structured overview of the landscape.

Quick Start & Requirements

• Installation: Primarily involves cloning the repository and navigating to individual paper code directories.
• Prerequisites: Specific Python libraries (e.g., TensorFlow, PyTorch), potentially CUDA for GPU acceleration, and often large datasets are required, varying per paper.
• Resources: Setup time and resource requirements are highly dependent on the specific paper's code and dataset.
• Links: The README provides links to individual paper code repositories and related datasets.

Highlighted Details

• Comprehensive categorization of DL applications in wireless communication.
• Daily updates via automated GitHub Actions for eess.SP and cs.IT arXiv categories.
• Includes links to datasets and researcher homepages for deeper engagement.
• Focus on reproducible research by linking papers to their code implementations.

Maintenance & Community

The project is community-driven, with contributions welcomed via GitHub issues. Key contributors are listed, and the project was initially released in February 2019, with a significant update in April 2021 for categorization and link completion.

Licensing & Compatibility

The repository itself does not specify a license. Individual code repositories linked within the project will have their own licenses, which may vary and could include restrictions on commercial use or closed-source linking.

Limitations & Caveats

The list is not exhaustive due to the author's limited focus and energy. While the project aims to include code, some papers may be listed without accompanying code, and the "TODO" section indicates planned additions like traditional communication papers and highly cited "Communication+DL" papers without code.