Packet Design is now part of      Learn More

Packet Design CTO to Discuss Self-Driving Network Customer Use Cases at AI Net and MPLS + SDN + NFV World Congress

Cengiz Alaettinoglu to describe how analytics and automation enable self-optimizing, self-healing networks, using real-world examples  

PARIS – April 11, 2017 – Advances in management software are making self-driving networks possible. Packet Design CTO Cengiz Alaettinoglu will discuss real use cases in Paris this week at both the MPLS + SDN + NFV World Congress and concurrent AI Net conferences.

Topic: Proactive Service Restoration and the Role of Machine Learning

In one case study Alaettinoglu will describe how, in response to the threat of an optical interface failure, critical traffic was rerouted to other paths in the network before the link failed. This was accomplished in a real-world environment with software that used link interface statistics and NetFlow data as well as routing telemetry and analytics to predict the interface issue. He will then explain how this self-driving network used machine learning algorithms for simple baselining as well as for complex path calculations using neural networks.

In addition, Alaettinoglu will provide other automation examples, including the provisioning of service paths with assured bandwidth or specific constraints, on-demand traffic engineering, and multi-layer traffic engineering.

Date: Thursday, April 12


Location: Marriott Conference Center and Hotel Paris

Packet Design is showing the latest versions of its Explorer Suite and Explorer SDN Platform in Booth 402 at the MPLS + SDN + NFV World Congress.

Earlier this week, Packet Design held its annual Customer Symposium, drawing a record number of attendees from four continents. Three customers delivered presentations about how they use the Explorer products. Industry partner Ciena, Blue Planet Division, gave the keynote about multi-layer orchestration, and Packet Design staff presented the company’s 2018 product roadmap, solution updates, and training sessions.