Wednesday, 14 March, 15:30-17:00
Theater II, Hall E
These companies have worked with machine learning to manage a number of applications in Optical networks. They are here to tell you about those experiences. Much more than a mere tutorial on how to get things started but real examples of operational efficiencies delivered. They will be talking about experiences they have had implementing a model perfecting it and then using it in their operations and the results they have achieved so far in the areas of open saving and people resource saving and other efficiency gains.
, Optics Consulting Engineer, Nokia, USA
Steve Plote, is Optics Consulting Engineer at Nokia. He is currently responsible for the support of all Nokia sales teams in the Americas as well as Channel Partners. Focusing on Network designs for the delivery of real time, next generation services for Video transport, Carrier Ethernet, Consumer Content Distribution, and Cloud Computing. He has personal responsibility for the network engineering and support for the WEB2.0 and Content Service Providers in North America. Mr. Plote has more than 30 years of experience in Data Center Interconnect, Telecommunications and LAN switching and transmission solutions. Prior to joining Nokia, he was Solutions Business Development and CSP Systems Engineering at BTI Systems and prior to that was Solutions Sales Director at Tellabs. He has many professional memberships and committee involvements including NANOG PC, OFC, MEF, IEEE, OSA.
Applications of Machine Learning in Facebook’s Production Network
Satyajeet Singh Ahuja, Network Architect, Facebook, USA
Satyajeet Singh Ahuja currently serves as Network Architect for Facebook. Previous positions include Network Architect at Google and System Architect at Infinera. His areas of interest include multi-layer network design, planning, modeling, risk assessment, and optimization.
Marginless Optical Networking with Machine Learning
, Research Engineer, Nokia, France
Yvan Pointurier received a Ph.D. from the University of Virginia, USA in 2006. Between 2006 and 2009, he was a postdoctoral fellow at McGill University in Montreal and then a senior researcher at AIT, Greece. In 2009, Dr. Pointurier joined Alcatel-Lucent (now Nokia) Bell Labs where he is now the head of the “Dynamic Optical Networking and Switching” department. His team is working on circuit and optical packet switched networks, with activities ranging from the physical layer to planning algorithms. Dr. Pointurier has authored or co-authored more than 15 European and US patents, and over 90 technical papers.
Practical applications of Machine Learning in Network Design, Optimization and Analysis: Opportunities and Challenges
, Network System Architecture, Google Inc., USA
Anurag Sharma is currently with the Network Architecture team at Google, focusing on the design, optimization, planning and analysis of Google’s cross-layer backbone network. Prior to joining Google, Anurag was the Sr. Principal SDN Architect at Infinera where he was responsible for the architecture of Infinera’s SDN product, Multi-Layer Path Computation Element (PCE) and North Bound API definitions, in addition to leading various open source networking efforts at IETF, ONF, and OIF.
Machine Learning for BER Estimation and Soft-Failure Mode Identification in Optical Transport Networks
Massimo Tornatore, Associate Professor, Politecnico di Milano, Italy
Massimo Tornatore is an Associate Professor with the Department of Electronics, Information, and Bioengineering, at Politecnico di Milano. He also holds an appointment as Adjunct Full Professor in the Department of Computer Science, University of California, Davis. His research interests include performance evaluation, optimization and design of communication and cloud networks (with an emphasis on the application of optical-networking technologies). He currently serves as an Editor for Photonic Network Communications, Optical Switching and Networking and IEEE Communication Surveys and Tutorials. He is a co-author of more than 280 scientific publications and was the co-recipient of ten best-paper awards.