Facial reconstruction

Search LJMU Research Online

Browse Repository | Browse E-Theses

Wireless Powered Mobile Edge Computing Systems: Simultaneous Time Allocation and Offloading Policies

Irshad, A, Abbas, ZH, Ali, Z, Abbas, G, Baker, T and Al-Jumeily, D (2021) Wireless Powered Mobile Edge Computing Systems: Simultaneous Time Allocation and Offloading Policies. Electronics, 10 (8). ISSN 2079-9292

[img]
Preview
Text
electronics-10-00965-v2.pdf - Accepted Version
Available under License Creative Commons Attribution.

Download (369kB) | Preview

Abstract

To improve the computational power and limited battery capacity of mobile devices (MDs), wireless powered mobile edge computing (MEC) systems are gaining much importance. In this paper, we consider a wireless powered MEC system composed of one MD and a hybrid access point (HAP) attached to MEC. Our objective is to achieve a joint time allocation and offloading policy simultaneously. We propose a cost function that considers both the energy consumption and the time delay of an MD. The proposed algorithm, joint time allocation and offload policy (JTAOP), is used to train a neural network for reducing the complexity of our algorithm that depends on the resolution of time and the number of components in a task. The numerical results are compared with three benchmark schemes, namely, total local computation, total offloading and partial offloading. Simulations show that the proposed algorithm performs better in producing the minimum cost and energy consumption as compared to the considered benchmark schemes.

Item Type: Article
Uncontrolled Keywords: 0906 Electrical and Electronic Engineering
Subjects: Q Science > QA Mathematics > QA75 Electronic computers. Computer science
T Technology > TK Electrical engineering. Electronics. Nuclear engineering
Divisions: Computer Science & Mathematics
Publisher: MDPI
Date Deposited: 23 Apr 2021 07:36
Last Modified: 23 Apr 2021 07:36
DOI or Identification number: 10.3390/electronics10080965
URI: https://researchonline.ljmu.ac.uk/id/eprint/14843

Actions (login required)

View Item View Item