Affiliation:
1. School of Information and Communication Engineering, Beijing University of Posts and Telecommunications, Beijing 100876, China
Abstract
Computation offloading has received a significant amount of attention in recent years, with many researchers proposing joint offloading decision and resource allocation schemes. However, although existing delay minimization schemes achieve minimum delay costs, they do so at the cost of losing possible further maximization of the number of serviced requests. Furthermore, the asymmetry between uplink and downlink poses challenges to resource allocation in edge computing. This paper addresses this issue by formulating the joint computation offloading and edge resource allocation problem as a mixed-integer nonlinear programming (MINLP) problem in an edge-enabled asymmetrical network. Leveraging the margin between a delay-minimum scheme and a near-deadline scheme, a water pumping and refilling (WPR) algorithm is proposed to maximize the number of accepted requests. The WPR algorithm can function both as a supplementary algorithm to a given offloading scheme and as a standalone algorithm to obtain a resource allocation scheme following a customizable refilling policy. The simulation results demonstrated that the proposed algorithm outperforms delay-minimum schemes in achieving a high acceptance ratio.
Subject
Physics and Astronomy (miscellaneous),General Mathematics,Chemistry (miscellaneous),Computer Science (miscellaneous)
Reference42 articles.
1. Joint computation offloading and user association in multi-task mobile edge computing;Dai;IEEE Trans. Veh. Technol.,2018
2. Mobile edge computing—A key technology towards 5G;Hu;ETSI White Pap.,2015
3. Computation Offloading Toward Edge Computing;Lin;Proc. IEEE,2019
4. Internet of Things (IoT) security with blockchain technology: A state-of-the-art review;Khan;IEEE Access,2022
5. Mobile-edge computing: Partial computation offloading using dynamic voltage scaling;Wang;IEEE Trans. Commun.,2016