Machine-Learning-Based Prediction and Optimization of Mobile Metro-Core Networks

We propose a methodology to optimize the decisions of mobile metro-core network orchestration systems. We use machine-learning-based traffic prediction to dynamically provision resources in advance. Resource allocation and reconfigurations are calculated through a heuristic that combines reinforcement learning and mixed integer linear programming.

[1]  Didier Colle,et al.  Power consumption modeling in optical multilayer networks , 2012, Photonic Network Communications.

[2]  Guido Maier,et al.  Network Orchestration for Dynamic Network Slicing for Fixed and Mobile Vertical Services , 2018, 2018 Optical Fiber Communications Conference and Exposition (OFC).

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

[4]  Guido Maier,et al.  Matheuristic with machine-learning-based prediction for software-defined mobile metro-core networks , 2017, IEEE/OSA Journal of Optical Communications and Networking.

[5]  LuJiaheng,et al.  Big data challenge , 2013 .

[6]  Guido Maier,et al.  Deep Learning-Based Traffic Prediction for Network Optimization , 2018, 2018 20th International Conference on Transparent Optical Networks (ICTON).