PhD in Engineering: Artificial Intelligence and RF Waveforms for radar and 5G communication
For this project we are looking for an enthusiastic student to develop new algorithms and waveforms for applications in the radar and communication domains. The student will be supervised by Dr Francesco Fioranelli and Prof Muhammad Imran at the School of Engineering, University of Glasgow, with the support of Leonardo Airborne and Space Systems.
Frequency spectrum is increasingly becoming an expensive and rare commodity, as we have more and more mobile users and services that need high volume of multimedia data traffic. This creates a “crunching” effect, potentially detrimental for telecommunication applications (cap to the number of users, need for additional infrastructures, increased latency unsuitable for real-time multimedia services), and even more for radar applications (the narrower allocated frequency bandwidth reduces resolution to sense targets of interest, and restricts the possibility of using different parts of the spectrum for different radar tasks, for example detection vs tracking).
To address this problem, researchers have developed innovative waveforms that can enable simultaneous transmission/reception of data, while achieving acceptable radar performances. A scenario to think of is a platoon of autonomous vehicles or a swarm of drones, where each vehicle aims to exchage data with the others (communication function) and at the same time measure distance and velocity of others (radar functions).
So, several examples of waveforms for co-existence and co-design of radar and communication do exist in the literature, but there are outstanding challenges: how do we select these waveforms, how do we decide which one is best depending on the scenario under test, and in terms of what metrics (mutual interference, ambiguity function, resolution in range and Doppler, bit error rate, …)? And also, can our system be intelligent and make these decisions on its own by looking at the environment and predicting its evolution? In this PhD project we will explore how machine learning techniques (e.g. reinforcement learning) can enable the self-configuration capability of intelligent systems for radar and communication, to address the aforementioned questions.
The candidate should have a degree in Engineering, Physics, or Mathematics, with knowledge of signal processing, ideally applied to radar and communication domains, and of programming and simulation languages (e.g. MATLAB, Simulink). Above all, the candidate should be willing to work on a novel exciting topic at the forefront of the technology. Interested candidates are invited to contact the first supervisor, Dr Fioranelli, to discuss their interests before applying.
The student will benefit from a close partnership with Leonardo Airborne and Space Systems. They will provide the candidate with training and company material, bimonthly reviews and advice from the industrial supervisor, annual presentation to the company radar branch of the business, and possible placements at the Leonardo site to implement and test the developed algorithms and waveforms. Leonardo is also sponsoring another PhD student in the Wireless Communications and Radar Group, plus a larger cohort of students across different universities allowing for mutual knowledge exchange and peer-learning for the students.
Funding is available to cover tuition fees for Home/EU applicants for 4 years, as well as paying a stipend at the Research Council rate (£14,777 for Session 2017-18).