Maray, Mohammed Mufareh A. (2021) Efficient computational offloading of dependent tasks in mobile edge networks. PhD thesis, University of Warwick.

PDF WRAP_Theses_Maray_2021.pdf - Submitted Version Embargoed item. Restricted access to Repository staff only until 21 December 2023. Contact author directly, specifying your specific needs. - Requires a PDF viewer. Download (33Mb)

Request Changes to record.

Abstract

Mobile Edge Network (MEN) is emerging as a novel computing paradigm that puts high storage and computational power within easy reach of mobile users for a range of applications such as big data applications and location-based services. The MENs consist of a number of small base stations, which we call Cloudlets, that provide the required services to end-users. Ecosystems are resource-constrained, making execution of resource-hungry applications challenging. Computation offload between ecosystems and cloudlets plays a key role in this vision and ensures that the integration between ecosystem and cloudlet is seamless with better quality of service such as lower latency. Analysis of the available literature relating to currently proposed offloading techniques focuses on centralised approaches with a small number of mostly static user devices hosting in-dependent tasks. In this thesis, we address three major offloading problems: (i) that algorithms consider distributed environment with multi offloading systems, (ii) users with ecosystems are mobile and (iii) tasks are dependent as (DAGs). We develop the offloading algorithms for mobile user devices with hosting of dependent tasks, where a dependent task cannot start until its immediate predecessor tasks have completed, with the aim of reducing completion latency. We start by formalizing the dependent task offloading problem as a constraint satisfaction problem with all proposed algorithms. While the first objective aims at minimising completion latency with central server in edge, the second objective aims at minimising completion latency with multi systems in distributed environment. We construct optimisation models for both objectives and develop two offloading algorithms to approximate the optimal solution. According to the results with ns-3 simulation, Optimisation CPLEX, and real deployment, our offloading algorithms are able to efficiently produce allocation schemes that are close to optimal during the offloading.

Item Type:	Thesis or Dissertation (PhD)
Subjects:	Q Science > QA Mathematics > QA76 Electronic computers. Computer science. Computer software
Library of Congress Subject Headings (LCSH):	Edge computing, Mobile computing, Cloud computing, Workflow -- Management -- Computer programs
Official Date:	December 2021
Dates:	Date Event December 2021 UNSPECIFIED
Institution:	University of Warwick
Theses Department:	Department of Computer Science
Thesis Type:	PhD
Publication Status:	Unpublished
Supervisor(s)/Advisor:	Jhumka, Arshad, 1974-
Sponsors:	Jāmiʻat al-Malik Khālid ; Saudi Arabia (Great Britain). Embassy
Format of File:	pdf
Extent:	xiii, 122 leaves : illustrations
Language:	eng

Request changes or add full text files to a record

Repository staff actions (login required)

View Item