Reconfigurable intelligent surface and UAV-assisted communications: a deep reinforcement learning approach

This thesis proposes novel methods based on the deep reinforcement learning algorithms (DRL) for maximising the energy efficiency (EE), sum-rate in reconfigurable intelligent surface (RIS) and unmanned aerieal vehicles (UAV)-aided wireless communications. The thesis carries out comprehensive optimization and evaluation of various DRL algorithms for several real-life applications including UAV's trajectory design, power allocation, data collection, wireless power transfer and RIS's phase shift matrix adjustment.<br/><br/>The thesis presents three major contributions. Firstly, we design a new UAV-assisted Internet-of -things (IoT) system relying on the shortest flight path of the UAVs while maximising the amount of data collected from IoT devices. then, a DRL-based technique is conceived for finding

Name dc:type.qualificationname

Doctor of Philosophy

Level dc:type.qualificationlevel

Doctoral Thesis

Grantor dc:publisher.institution

Queen's University Belfast

Year dc:date.issued

2022

Author dc:creator

• Nguyen, Khoi Khac

Advisor dc:contributor.advisor

• Duong, Quang

Language dc:language

eng

Identifier

oai:pure.qub.ac.uk/portal:studenttheses/d4dc3bc9-321a-43e3-9849-bc87f4c9b1ce

OAI identifier oai:identifier

oai:pure.qub.ac.uk/portal:studenttheses/d4dc3bc9-321a-43e3-9849-bc87f4c9b1ce