Scaling Graph-based Deep Learning models to larger networks

Graph Neural Networks (GNN) have shown a strong potential to be integrated into commercial products for network control and management. Early works using GNN have demonstrated an unprecedented capability to learn from different network characteristics that are fundamentally represented as graphs, such as the topology, the routing configuration, or the traffic that flows along a series of nodes in the network. In contrast to previous solutions based on Machine Learning (ML), GNN enables to produce accurate predictions even in other networks unseen during the training phase. Nowadays, GNN is a hot topic in the Machine Learning field and, as such, we are witnessing great efforts to leverage its potential in many different fields (e.g., chemistry, physics, social networks). In this context, the Graph Neural Networking challenge 2021 brings a practical limitation of existing GNN-based solutions for networking: the lack of generalization to larger networks. This paper approaches the scalability problem by presenting a GNN-based solution that can effectively scale to larger networks including higher link capacities and aggregated traffic on links.

[1]  Razvan Pascanu,et al.  Discovering objects and their relations from entangled scene representations , 2017, ICLR.

[2]  Kouichi Sakurai,et al.  One Pixel Attack for Fooling Deep Neural Networks , 2017, IEEE Transactions on Evolutionary Computation.

[3]  Fabien Geyer,et al.  DeepTMA: Predicting Effective Contention Models for Network Calculus using Graph Neural Networks , 2019, IEEE INFOCOM 2019 - IEEE Conference on Computer Communications.

[4]  Razvan Pascanu,et al.  Relational inductive biases, deep learning, and graph networks , 2018, ArXiv.

[5]  Matthew Roughan,et al.  The Internet Topology Zoo , 2011, IEEE Journal on Selected Areas in Communications.

[6]  Roberto Proietti,et al.  Deep-RMSA: A Deep-Reinforcement-Learning Routing, Modulation and Spectrum Assignment Agent for Elastic Optical Networks , 2018, 2018 Optical Fiber Communications Conference and Exposition (OFC).

[7]  Christopher R. Palmer,et al.  Generating network topologies that obey power laws , 2000, Globecom '00 - IEEE. Global Telecommunications Conference. Conference Record (Cat. No.00CH37137).

[8]  Razvan Pascanu,et al.  Interaction Networks for Learning about Objects, Relations and Physics , 2016, NIPS.

[9]  Aleksander Madry,et al.  Exploring the Landscape of Spatial Robustness , 2017, ICML.

[10]  Ah Chung Tsoi,et al.  The Graph Neural Network Model , 2009, IEEE Transactions on Neural Networks.

[11]  Yoshua Bengio,et al.  Empirical Evaluation of Gated Recurrent Neural Networks on Sequence Modeling , 2014, ArXiv.

[12]  Zhiyuan Liu,et al.  Graph Neural Networks: A Review of Methods and Applications , 2018, AI Open.

[13]  Samuel S. Schoenholz,et al.  Neural Message Passing for Quantum Chemistry , 2017, ICML.

[14]  George Varghese,et al.  Efficient fair queueing using deficit round-robin , 1996, TNET.

[15]  Dafna Shahaf,et al.  Learning to Route , 2017, HotNets.

[16]  John A. Wegener,et al.  The Graph Neural Networking Challenge:A Worldwide Competition for Education in AI/ML for Networks , 2021 .

[17]  Yusheng Ji,et al.  Understanding the Modeling of Computer Network Delays using Neural Networks , 2018, Big-DAMA@SIGCOMM.

[18]  Xin Wang,et al.  Machine Learning for Networking: Workflow, Advances and Opportunities , 2017, IEEE Network.

[19]  Albert Cabellos-Aparicio,et al.  Unveiling the potential of Graph Neural Networks for network modeling and optimization in SDN , 2019, SOSR.

[20]  Zhibo Gong,et al.  Deep-Q: Traffic-driven QoS Inference using Deep Generative Network , 2018, NetAI@SIGCOMM.