Polina Bayvel
Events
Polina Bayvel
University College London
What Will You Use Those Extra GPUs For? Designing Scalable Optical Networks for an AI-Driven World
To support growing data demands, partly driven by AI applications, optical networks must deliver massive capacity with intelligence and efficiency. However, optical networks are not just sets of transparent pipes, they have physical transmission and graph properties which must be integrated into the network design – both for new networks and to evolve existing network infrastructure. Optimising over tens of formats, hundreds of independent channels over thousands of kms through brute force optimisation is hard, if not impossible! Reduction of complexity is key. By integrating advanced optimisation and machine learning, we must learn to design that match the complexity of future applications and the talk will look at some possible direction to achieve this.
About the Speaker
Professor Polina Bayvel is the Head of the Optical Networks Group at UCL, which she founded in 1994. Her research focuses on optical communications and networks, including intelligent optical networks, wavelength routing, high-speed transmission and fiber nonlinearity mitigation.
After completing her PhD, she worked as a systems engineer at STC Submarine Systems (now Alcatel) and Nortel, specializing in optical transmission and network planning. In 1994, she received a Royal Society University Research Fellowship and established the first academic systems engineering group in optical networks at UCL.
A Fellow of the Royal Society, Royal Academy of Engineering, IEEE, and Optica, she was awarded a CBE in 2017 for services to engineering. She is the first woman to receive the Thomas Young Medal (2021) and the Royal Society Rumford Medal (2023). In 2024, she was honored with the Humboldt Research Prize.
Bayvel has authored over 500 journal and conference papers, led the EPSRC Programme Grant UNLOC (2012-2018) and currently leads the EPSRC Programme Grant TRANSNET (2018-2024), which aims to revolutionize optical networks using machine learning and intelligent transceivers. She advocates for secure, low-delay, high-capacity communications infrastructure to support the digital economy and transformative technologies.
