Deep Reinforcement Learning (DRL) algorithms have been recently proposed to solve dynamic Radio Resource Management (RRM) problems in 5G networks. However, the slow convergence experienced by traditional DRL agents puts many doubts on their practical adoption in cellular networks. In this paper, we first discuss the need to have accelerated DRL algorithms. We then analyze the exploration behavior of various state-of-the-art DRL algorithms for slice resource allocation, and compare it with the traditional 5G Radio Access Network (RAN) slicing baselines. Finally, we propose a transfer learningaccelerated DRL-based solution for slice resource allocation. In particular, we tackle the challenge of slow convergence by transferring the policy learned by a DRL agent at an expert base station (BS) to newly deployed agents at target learner BSs. Our approach shows a remarkable reduction in convergence time and a significant performance improvement compared with its nonaccelerated counterparts when tested against multiple traffic load variations.
With the advancements in wireless network technologies over the past few decades and the deployment of 4G LTE networks, the capabilities and services provided to end-users have become seemingly endless. Users of smartphones utilize high-speed network services while commuting on public buses and hope to have a consistent, high-quality connection for the duration of their trip. Due to the massive load demand on cellular networks and frequent changes in the underlying radio channel, users often experience sudden unexpected variations in the connection quality. To overcome such a variation and maintain a consistent connection, we need to predict these variations before they occur. This can be accomplished by analyzing different network quality parameters at various times and locations and investigating the main factors that affect the network's performance and network QoS. To this end, we conducted a network survey via Kingston Transit, in Kingston, Ontario, using the Android network monitoring application G-NetTrack Pro from which we constructed a dataset of various client-side wireless network quality parameters. The dataset consists of 30 repeated public transit bus trips, each lasting no more than one hour. We studied two techniques for throughput analysis: regression predictive modelling and time series forecasting. For regression predictive modelling, we deployed various machine learning models on the collected data for throughput prediction and achieved the highest prediction performance with the random forest model. For time series forecasting, we used statistical methods as well as deep learning architectures. Our evaluation shows that the machine learning models had a higher throughput prediction performance than the time series forecasting techniques. In this thesis, we present an analysis of the collected data, where we investigate the effects of time and location on the network's measured throughput and signal strength. Also, we discuss and compare the results of applying different throughput prediction techniques on the collected data. iii Acknowledgements I would like to express my sincere gratitude to my supervisor Prof. Hossam Hassanein for his endless guidance, support and constant feedback throughout this research work. Also, I would like to extend my appreciation to my co-supervisor Prof. Aboelmagd Noureldin for his valuable suggestions and insights during the planning and development of this research work. I am deeply grateful to Prof. Hazem Abbas for his guidance and motivation, and for giving his time so generously to help me complete this research work. I also wish to thank Basia Palmer for her patient and accurate review of this thesis and her useful suggestions. I would like to thank all my colleagues in Queen's Telecommunications Research Lab. In particular, I would like to thank Ahmad Nagib for his generous help and support, Basma, Rawan, Sara and Mary for their continuous encouragement. To my parents, Ashraf and Amany, thank you for always believing in me in every step of my life. Words c...
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.