A Machine Learning Approach for Traffic Flow Provisioning in Software Defined Networks

Abstract

With the recent surge of machine learning and artificial intelligence, many research groups are applying these techniques to control, manage, and operate networks. Soft-ware Defined Networks (SDN) transform the distributed and hardware-centric legacy network into an integrated and dynamic network that provides a comprehensive solution for managing the network efficiently and effectively. The network-wide knowledge provided by SDN can be leveraged for efficient traffic routing in the network. In this work, we explore and illustrate the applicability of machine learning algorithms for selecting the least congested route for routing traffic in a SDN enabled network. The proposed method of route selection provides a list of possible routes based on the network statistics provided by the SDN controller dynamically. The proposed method is implemented and tested in Mininet using Ryu controller.