Enabling Power Efficient Optical Communication through Novel Digital Signal Processing (EPIC DSP)

Project ID: 2531ad1532

(You will need this ID for your application)

Research Theme: Information and Communication Technologies

UCL Lead department: Electronic and Electrical Engineering (EEE)

Project Summary:

The digital communication infrastructure underpinning the internet represents 3% of global power usage, and due to growing internet traffic, this is expected to increase ever further. Of this, optical communication systems contribute 30% to power usage. To address the energy consumption, we are investigating low-complexity digital signal processing (DSP), enabling the combined use of energy-efficient semiconductor optical amplifiers (SOAs) and hollow core optical fibre (HCF). With the aim to create new communication system architectures with significantly higher energy efficiency.

The use of SOAs, which have non-linear gain dynamics, introduces non-linear distortion that degrades data transmission performance in optical transmission systems. This studentship will explore the compensation of this distortion using machine learning algorithms and DSP together with accurate modelling of optical transmission systems. What are the theoretical limits on throughput and what is the relationship between throughput, complexity and power consumption of different approaches? Within this studentship you will develop the skills to design novel modelling, DSP and machine learning algorithms with the aim of demonstrating record energy efficiency using our world-leading testbed.

You will join the Optical Networks Group (ONG) at UCL, an internationally leading group in optical communications research. Our work spans all distance scales—from micrometre-scale photonic integrated circuits for ultra-fast data centres to 10,000-kilometer intercontinental fibre systems that form the backbone of the internet. You will collaborate with experts dedicated to ensuring that our global communications infrastructure meets the ever-growing demand for data.

We seek motivated candidates with a strong background in electrical engineering, physics, computer science, or a related discipline. If you have an interest in optical communications, DSP, and machine learning, we would love to hear from you. Join us to make a meaningful impact on the future of global communications and develop highly valued skills for academia and industry.