network-modeling-GNN

This repository includes a collection of Graph Neural Network (GNN) models applied to computer network modeling.

1. Towards more realistic network models based on Graph Neural Networks

This work extends RouteNet to include support for modeling end-to-end performance in networks with forwarding nodes of variable qeue size.

A. Badia-Sampera, J. Suárez-Varela, P. Almasan, K. Rusek, P. Barlet-Ros, A. Cabellos-Aparicio.

Proceedings of the 2019 ACM CoNEXT 2019 Student Workshop, Orlando, USA, December 2019.

@inproceedings{routenet-nodes,
  title={Towards more realistic network models based on Graph Neural Networks},
  author={Badia-Sampera, Arnau and Su\'arez-Varela, Jos\'e and Almasan, Paul and Rusek, Krzysztof and Barlet-Ros, Pere and Cabellos-Aparicio, Albert},
  booktitle={Proceedings of the 2019 ACM CoNEXT 2019 Student Workshop},
  pages={},
  year={2019}
}

Please find the implementation of the GNN model presented in this paper at the follwoing [link](routenet_with_forwarding_nodes).