Artificial Intelligence-Enabled Free-Space Optics for Future Advanced Satellite Communications

Given their significant contribution to various aspects of modern life, satellite communications have become an essential part of our world today. Significant efforts are being made globally to develop the next-generation satellite systems that can offer high capacity to support the massive data traffic from numerous applications such as climate change mitigation, disaster prevention, IoT, and beyond 5G/6G. Up to this point, space communications have primarily depended on Radio Frequency (RF) technology. NASA and ESA have recently conducted research activities that indicate Free-Space Optics (FSO) as a promising technology for future space communications. FSO offers several benefits over RF, such as higher data rates, increased security, and smaller, lighter terminals with lower power needs [1-4]. 

The practical implementation of FSO for space communications is not trivial and requires overcoming several challenges. Atmospheric turbulence is a result of random fluctuations in the atmosphere refractive index, which can cause fluctuations in the received power of optical links between satellites and ground stations. Optical beam alignment errors can occur due to satellite vibrations, resulting in a decrease in received power. Additionally, space radiations can interfere with signal by introducing noise, which can reduce the signal-to-noise ratio and result in communication errors. The performance of optical satellite communication systems is significantly affected by these factors [5].

Due to the unpredictable nature of the FSO channel impairments, traditional communication mitigation techniques have been found to be inadequate in addressing the issues. In response, artificial intelligence (AI) is increasingly viewed as an innovative and promising solution to address the challenges.

In this project, we combine free-space optics and artificial intelligence technologies to address key challenges in current satellite communications. This project aims to carry out world-leading research on building a cutting-edge FSO/AI-based communication platform to accelerate intelligent, reliable, secure, and high-capacity satellite communications. FSO will be used for high-speed and secure data transmissions between ground stations and satellites or between satellites. To enhance the reliability of optical satellite communications, AI techniques will be studied for one or more of following objectives.

• AI-based predictive models will be explored to predict atmospheric turbulence channel conditions, enabling the adjustment of transmission parameters to counteract the effect of atmospheric turbulence.
• AI-based pattern recognition algorithms will be researched to identify and learn patterns of alignment errors, with the aim of optimizing the optical beam control to compensate for these errors.
• AI-based optimization algorithms will be studied to improve and adapt several receiver components, such as optical bandpass filters and amplifiers, to mitigate the impact of space radiations on the system performance.

Combining free-space optics and artificial intelligence technologies to develop an FSO/AI-based communication platform is a new and innovative concept in the field of satellite communication. This research has the potential to significantly advance satellite communications by improving reliability, security, intelligence, and capacity, while keeping low size, weight, and power. It could lead to the development of a new generation of satellite communications to meet demands wide range of applications from climate change mitigation to 5G/6G.

The proposed research will take place at the Centre for Space AI and School of Computer Science, Artificial Intelligence and Electronics. We have a team of experts in FSO, satellite communications and AI, who will assist with the development and testing of the proposed FSO/AI-based communication platform. The centre also provides access to state-of-the-art facilities, such as computer labs, electronics labs, test chambers, and measurement facilities, for both development and testing purposes.

The project will involve utilizing methodologies related to communication channel modelling, signal processing, cross-layer design, and AI/ML. Applicants should have research experience or a strong willingness to develop knowledge and research skills in these areas. Students with a background in electrical and electronics engineering, telecommunications engineering, computer science, physics, or mathematics are particularly encouraged to apply.

How to apply

Formal applications can be made via the University of Bradford web site. Please select 'Full-time PhD in Electrical Engineering' as the course.