Graph neural networking challenge 2021: Creating a scalable network digital twin

Go back to programme

Graph neural networking challenge 2021: Creating a scalable network digital twin

  • Watch on Zoom

    Watch on YouTube

    * Register (or log in) to the AI4G Neural Network to add this session to your agenda or watch the replay

  • Graph Neural Networks (GNN) have produced groundbreaking applications in different fields where data is fundamentally structured as graphs (e.g., chemistry, recommender systems). In the field of computer networks, this new type of neural networks is being rapidly adopted for a variety of relevant networking use cases, particularly for those involving large and complex graphs (e.g., performance modeling, routing optimization, resource allocation in wireless networks).

    This talk presents the “Graph Neural Networking challenge 2021”. The second edition of this competition brings a fundamental limitation of existing GNNs: their lack of generalization capability to larger graphs. In order to achieve production-ready GNN-based solutions, we need models that can be trained in network testbeds of limited size, and then be able to operate with guarantees in any real customer network, which are often 10x larger in number of nodes. In this challenge, participants are asked to train their GNN models in small network scenarios (up to 50 nodes), and then test their accuracy in networks of increasing size not seen before, up to 300 nodes. Solutions with better scalability properties will be the winners.

    • i

      Speaker abstract

      Forging the Networks of Tomorrow: Leveraging Deep Learning for Mobile Network planning and operation - The networking community is actively engaged in the search for the key technologies that will drive the success of 6G networks. In this exciting landscape, Deep Learning can be a game-changer in propelling such a revolution, especially for processing the vast amounts of data collected in networks, uncovering intricate patterns in that data, and making complex decisions in real time. This talk will present some opportunities and ongoing efforts to apply Deep Learning for planning and operation in mobile networks. We will introduce some use cases where we leverage Deep Learning for achieving unprecedented levels of connectivity and user experience. Next, we will discuss some key open challenges to achieve mature Deep Learning solutions for networks, with a particular focus on energy efficiency. Finally, we will outline some future research directions that may help materialize Deep Learning-based solutions for the Mobile Networks of Tomorrow.
      José Suárez-Varela
      Telefonica Research
    Share this session

    Are you sure you want to remove this speaker?