OpenDaylight Summit 2016

In recent years, Machine Learning in networking and SDN have attracted interests from both the academia and the industry. Most solutions ship the data out of the SDN controller and analyze them using specialized tools. In this presentation, we present a solution that creates a Machine Learning framework (MILA) on top of the leading SDN controller OpenDaylight (ODL). By providing machine learning capabilities directly from the controller, the framework simplifies the intelligent SDN applications creation and enables the controller to continuously learn from the data generated in the network.

In this presentation, we will go through typical machine learning use cases in SDN, explain the scalable MILA architectural design, and present a demo that leverages the framework and ODL Time Series Data Repository to apply Machine Learning algorithms and generate the valuable data insights.