A New Offloading Technique Based on Deep Learning for Mobile-Edge Computing

Shetty, Prathista Santhosh Kumar (2022) A New Offloading Technique Based on Deep Learning for Mobile-Edge Computing. Masters thesis, Dublin, National College of Ireland.

Preview	PDF (Master of Science) Download (802kB) | Preview
Preview	PDF (Configuration manual) Download (617kB) | Preview

Abstract

Offloading techniques have gained more importance in today's world because of its need for high utilization of computing power, storage capacity and energy. The fundamental principle is to transfer resource-intensive processing operations to a more powerful processor for rapid computation. The development of new mobile technologies is increasing, requiring efficient utilization of resources. Mobile Edge Computing with distributed network has been developed with an aim to bring all the computational servers close to the end-users that meets predefined requirements. The rate at which tasks arrive at the mobile edge computing server for computing fluctuates every minute and it also depends on the user’s density. For optimal computational rate in such a dynamic environment, task offloading should be efficient to reduce the uploading and downloading time. The aim of this research is to offload the complexity brought on by the rapid advancements of distributed computing in Mobile Edge technologies. To address this issue, a framework that employs a scalable solution using deep neural network solution is used that trains itself based on the binary offloading predictions by using the K-modes algorithm that chooses the largest reward. Since combinatorial optimization issues are no longer need to be solved, its computational complexity is significantly lesser than 0.1 % on a large user network with 30 systems. The research also shows that the CPU latency is about 67 times lesser when compared to the coordinate descent approach.

Item Type:	Thesis (Masters)
Supervisors:	Name Email Sahni, Vikas UNSPECIFIED
Subjects:	Q Science > QA Mathematics > Electronic computers. Computer science T Technology > T Technology (General) > Information Technology > Electronic computers. Computer science T Technology > T Technology (General) > Information Technology > Cloud computing Q Science > Q Science (General) > Self-organizing systems. Conscious automata > Machine learning
Divisions:	School of Computing > Master of Science in Cloud Computing
Depositing User:	Tamara Malone
Date Deposited:	19 Apr 2023 15:07
Last Modified:	19 Apr 2023 15:07
URI:	https://norma.ncirl.ie/id/eprint/6492

Actions (login required)

View Item