jwwthu/GNN-Communication-Networks

A curated, actively maintained collection of academic papers and resources about applying graph-based deep learning to communication networks.

What is this?

This GitHub repository is a reading list and resource hub. It gathers research papers, surveys, competitions, and tools related to using graph-based deep learning (a type of AI that works on data structured as connected nodes and edges) in the field of communication networks — such as wireless systems, routing, network security, and edge computing.

Who maintains it?

The repository appears to be run by an author (likely Jianping Wang) who has personally published multiple peer-reviewed works in this area, including a survey paper and studies on federated learning and routing optimization. The maintainer also runs a Chinese-language social media account (WeChat and Zhihu) called “网络与通信” (Network and Communication), which periodically posts conference deadlines and summaries of open-source papers.

What's inside?

• Literature list — The core of the repo. A large, monthly-updated bibliography of recent papers sorted by year (currently 2025 and 2026) and by type (journal articles vs. conference papers). Each entry has a title, authors, and a link to the original publication.
• Other surveys — Links to external survey papers by other research groups, covering topics like graph neural networks in wireless networks, IoT, and resource allocation.
• Competition — References to a worldwide competition on graph neural networking.
• Tool — A link to a prototyping tool called IGNNITION, designed to help build graph neural networks for communication systems.
• survey.md — A separate file that the maintainer points to as a companion to their own survey paper, listing studied use cases.

How is it organized?

The README functions as a table of contents. Sections are separated by topic (surveys, competition, tool, literature), and the literature section is further split by year and publication venue type. There's no code repository here — it's purely a curated index of external resources.

The list is described as “updated monthly,” so it's intended to be a living, evolving collection rather than a static document.

Who is this for?

Researchers, graduate students, or engineers in networking or wireless communications who want a running, hand-curated bibliography of recent AI-for-networks papers without searching through multiple databases themselves.