Tuesday, December 4, 2018

MEF18 PoC - Machine Learning-Based Elastic Network Services



MEF18 Proof of Concept, 29 - 31 Oct - Machine Learning-Based Elastic Network Services. PoC Participants: Colt Technology Services & Accedian.

Speaker: Tom Foottit, Senior Product Manager, Accedian.

This PoC demonstrates the ability to leverage network performance data and machine learning to provide customer visibility to service performance and also demonstrates how to use current and historical service performance to dynamically adjust service parameters in a closed loop.


The PoC demonstrates this by using (1) Accedian and other network equipment to collect PM data, (2) the Accedian DATAHUB platform to ingest and run machine learning (ML) algorithms on the PM data, (3) the Colt Novitas system to provide performance visibility, and (4) the Colt orchestration layer to complete closed loop automation in the network using the output of ML from DATAHUB to adjust services based on current and historical network conditions. These systems are tied together using open APIs in alignment with the MEF LSO framework.

This showcase is a demonstration of the additional power that analytics and machine learning can bring to MEF LSO and the MEF 3.0 automated network.

Save the date for MEF19, 18-22 November 2019, JW Marriott, LA LIVE, Los Angeles.

https://youtu.be/Aji_c62Ej7g