Monday, March 9, 2020
2:00 PM -
Darko Zibar, Technical University of Denmark, DTU Fotonik, Denmark; Alan Pak Tao Lau; The Hong Kong Polytechnic University, Hong Kong; Jelena Pesic, Nokia Bell Labs, France
The field of machine learning provides a new set of tools to the optical communication community. These tools are valuable when the relationship between the input and the output of the system is highly complex or when there is a lack of analytical models describing the relation between the input and the output of the system.
An example of such a situation is when communicating over the fiber-optic channel, which is inherently nonlinear and highly complex. As such, optimal transmitter and receiver architectures that maximize the capacity—distance product of fiber optic channel are still unknown. Moreover, to satisfy future capacity demands, next generation optical communications may employ a combination of ultra-wideband transmission (O+S+E+C+L) and spatial-division multiplexing. Those systems will be highly complex, and bringing the tools from the machine learning community to design transmitter, and receiver architecture as well as optical amplification schemes may prove very beneficial. Finally, performing network monitoring and optimization for improved network efficiency is rather challenging due to many transmission parameters involved.
The goal of this symposium is to gather active groups working on this emerging topic and chart the future for the application of machine learning in optical communication systems and networks.
Keisuke Kojima; Mitsubishi Electric Research Laboratories, USA
Deep Learning for Inverse Design of Optical Device
Maxim Kuschnerov; Huawei, Germany
Advances in Deep Learning for Digital Signal Processing in Coherent Optical Modems
Josh Gordon; NIST, USA
Summary: NIST Workshop on Machine Learning for Optical Communication Systems
Cristina Rottondi; Politecnico di Torino, Italy
Active vs Transfer Learning for QoT Estimation with Low Number of Probes in Optical Networks
Andrew Shiner; Ciena, Canada
Neural Network Training for OSNR Estimation - from Prototype to Product
Åsa Ribbe; EPO, Germany
Towards Intelligent Optical Networks: The Role of Intellectual Property
Marija Furdek; Chalmers University of Technology, Sweden
Machine Learning for Optical Network Security Management