Optimising computational offloading and resource management in online and stochastic fog computing systems

Alenizi, Faten 2023. Optimising computational offloading and resource management in online and stochastic fog computing systems. PhD Thesis, Cardiff University. Item availability restricted.

	PDF (Faten Alenizi, PhD, Thesis) - Accepted Post-Print Version Restricted to Repository staff only until 8 December 2024 due to copyright restrictions. Available under License Creative Commons Attribution Non-commercial No Derivatives. Download (5MB)
	PDF (Cardiff University Electronic Publication Form) - Supplemental Material Restricted to Repository staff only Download (285kB)

Abstract

Fog computing is a potential solution to overcome the shortcomings of cloud-based processing of IoT tasks. These drawbacks can include high latency, location awareness, and security attributed to the distance between IoT devices and cloud-hosted servers. Although fog computing has evolved as a solution to address these challenges, it is known for having limited resources that need to be effectively utilised. This is because its advantages could be lost. Moreover, the increasing number of IoT devices and the amount of data they generate make optimising Quality of Service (QoS) in IoT applications, computational offloading, and managing fog resources more challenging. In this context, the problem of computational offloading and resource management is investigated in online and stochastic fog systems. To deal with dynamic online fog systems, we propose a combination of two algorithms: dynamic task scheduling (DTS) and dynamic energy control (DEC). These methods were applied with a fixed offloading threshold (i.e., the criteria by which a fog node decides whether tasks should be offloaded to a neighbour, and which neighbour, rather than executed locally) with the aim to minimise overall delay, improve the throughput of user tasks, and minimise energy consumption at the fog layer while maximising the use of resource-constrained fog nodes. The approach is further enhanced by applying a dynamic offloading threshold. Compared to other benchmarks, our approach could reduce latency by up to 95.4%, improve throughput by 41%, and reduce energy consumption by up to 55.7% in fog nodes. For stochastic fog systems, we address the computational offloading and resource management problem. This is with the aim to minimise the average energy consumption of fog nodes while meeting QoS requirements of tasks. We formulated the problem as a stochastic problem and decomposed it into two subproblems. In order to solve this problem, we have proposed a scheme called Joint Q-learning and Lyapunov Optimization (JQLLO). Using simulation results, we demonstrate that JQLLO outperforms a set of baselines.

Item Type:	Thesis (PhD)
Date Type:	Acceptance
Status:	Unpublished
Schools:	Computer Science & Informatics
Subjects:	Q Science > QA Mathematics > QA75 Electronic computers. Computer science Q Science > QA Mathematics > QA76 Computer software
Date of First Compliant Deposit:	8 December 2023
Date of Acceptance:	March 2023
Last Modified:	13 Dec 2023 13:24
URI:	https://orca.cardiff.ac.uk/id/eprint/164630

Actions (repository staff only)

Edit Item