Maharaj, Sunil2023-02-272023-02-272023-05-122023*A2023https://repository.up.ac.za/handle/2263/89854Thesis (PhD (Electronic Engineering))--University of Pretoria, 2023..The increased densification of wireless networks has led to the development of integrated access and backhaul (IAB) networks. In this thesis, deep reinforcement learning was applied to solve resource management and backhaul routing problems in millimeter-wave IAB networks. In the research work, a resource management solution that aims to avoid congestion for access users in an IAB network was proposed and implemented. The proposed solution applies deep reinforcement learning to learn an optimized policy that aims to achieve effective resource allocation whilst minimizing congestion and satisfying the user requirements. In addition, a deep reinforcement learning-based backhaul adaptation strategy that leverages a recursive discrete choice model was implemented in simulation. Simulation results where the proposed algorithms were compared with two baseline methods showed that the proposed scheme provides better throughput and delay performance.en© 2022 University of Pretoria. All rights reserved. The copyright in this work vests in the University of Pretoria. No part of this work may be reproduced or transmitted in any form or by any means, without the prior written permission of the University of Pretoria.UCTDWireless CommunicationsMachine learningCongestion controlDeep reinforcement learningIntegrated access and backhaulThesisu10410903https://doi.org/10.25403/UPresearchdata.22182295