SC483 - VIRTUAL: Machine Learning in Optical Networks
Monday, 07 March
08:30 - 12:30 (Pacific Time (US & Canada), UTC - 08:00)
Short Course Level:
Massimo Tornatore, Politecnico di Milano, Italy
Darko Zibar, DTU FOTONIK, Denmark
Short Course Materials
You must be logged in and registered for this short course to access the class and documents Log In Now.
Access to Short Course Materials
Access to the short course materials, including the link to join the event on zoom will be available 30 minutes prior to the scheduled start and up to 60 minutes after the scheduled end.
Short Course Description:
*This course is virtual only*
Machine learning (ML) has recently attracted a surge of interest in optical networking and communication research due to its pattern recognition and predictive capabilities for various key applications. Large-scale monitoring data are generated every day in optical networks, which makes ML a promising solution for decision making. In this short course, we introduce the fundamental concepts and principles of ML, with a special focus on multi-layer neural networks. We survey existing work on various applications in optical domain, ranging from predictive maintenance, quality of transmission estimation and component design. Finally, we provide a tutorial for participants showing how to implement backpropagation algorithm for learning neural network adaptive weights. We aim to provide a general overview of the key problems, common formulations, existing methodologies and future directions. This course will inspire the audience and facilitate ML research and development in optical networking and communication systems.
The outline is given below,
- Fundamental concepts of ML
- ML model learning and evaluation
- Neural Networks (NN) fundamentals
- ML Applications
- Quality of Transmission estimation
- Failure detection and identification
- Optical Amplifier Control
- Component Design
- Overview of other applications
- Tutorial on numerical methods for NN optimization
Short Course Benefits:
This course should enable participants to,
- Understand the main machine learning categories
- Become familiar with the most relevant ML algorithms used in practice (with a focus on neural networks)
- Understand the current status of the ML technology and its applications in optical networking and communications
- Compare properties and requirements for various ML evaluation techniques
- Gain insights on how to implement a proof-of concept algorithm for neural network learning
Short Course Audience:
The course is intended for interested people from academia and industry without any previous knowledge in machine learning. A basic understanding of optical fiber transmission and programming can be helpful but is not a hard requirement. Attendance is also beneficial for machine learning experts with limited optical networking background who want to learn about the potential applications in the area of optical communication and networking.
Massimo Tornatore is an Associate Professor at Politecnico di Milano, Italy. He also holds an appointment as Adjunct Professor at University of California, Davis, USA and as visiting professor at University of Waterloo, Canada. His research interests include performance evaluation and design of communication networks (with an emphasis on optical networking), and machine learning application for network management. He co-authored more than 400 conference and journal papers (with 19 best-paper awards) and of the recent Springer “Handbook of Optical Networks”. He is member of the Editorial Board of IEEE Communication Surveys and Tutorials, IEEE Communication Letters, IEEE Transactions on Network and Service Management and Elsevier Optical Switching and Networking.
Darko Zibar Darko Zibar is Professor at the Department of Photonics Engineering, Technical University of Denmark and the group leader of Machine Learning in Photonics Systems (M-LiPS) group. He received M.Sc. degree in telecommunication and the Ph.D. degree in optical communications from the Technical University of Denmark, in 2004 and 2007, respectively. Some of his major scientific contributions include: record capacity hybrid optical-wireless link (2011), record sensitive optical phase noise measurement technique that approaches the quantum limit (2019) and design of ultrawide band arbitrary gain Raman amplifier (2019). He is a recipient of Young Researcher Award by University of Erlangen-Nurnberg (2016) and European Research Council (ERC) Consolidator Grant (2017). Finally, he was a part of the team that won the HORIZON 2020 prize for breaking the optical transmission barriers (2016).