Resource Management in Container-based Mobile Edge Computing
Loading...
Journal Title
Journal ISSN
Volume Title
Sähkötekniikan korkeakoulu |
Master's thesis
Unless otherwise stated, all rights belong to the author. You may download, display and print this publication for Your own personal use. Commercial use is prohibited.
Author
Date
2018-10-08
Department
Major/Subject
Communications Engineering
Mcode
ELEC3029
Degree programme
CCIS - Master’s Programme in Computer, Communication and Information Sciences (TS2013)
Language
en
Pages
vii+67
Series
Abstract
Mobile edge computing is a promising technology which provides support to time-sensitive applications by pushing centralized cloud processing capabilities to distributed Fog nodes. These fog nodes are deployed at one-hop distance from end-user and provide real-time data processing capabilities at the edge of network. Due to service provisioning at the edge of network, no congestion occurs at the core of network, quality of service (QoS) is improved and the overall network operational cost is significantly reduced. However, these nodes have limited capabilities such as processing, storage and coverage so, they face challenge of mobility support for a mobile user when continued service (i.e. zero downtime) is required during handovers between edge nodes. Furthermore, they also need an effective task allocation and resource management strategy to ensure smooth operation of edge services. Unlike traditional VM based environment in Fog Computing, this work explores lightweight Docker containers to deploy and migrate services. In this work, an interactive event-driven dashboard is developed for real-time edge node registration, system monitoring, service initiation and migration. Then, motivated by Fog Following Me, a couple of resource allocation schemes (i.e. algorithm-I & II) have been introduced to dynamically manage the compute resources among fog nodes. For smooth service operation and stable migration, an application profiling feature has been introduced which assigns the needed quota for an application requirement in terms of CPU, GPU and RAM. The developed system's performance is evaluated by conducting various experiments. The experimental results clearly demonstrate and verify the working feasibility of the whole system's operation in context of edge computing. However, the observed processing delays during service migration marks the limitation of Docker and suggest the need to use latest optimization tools to cut down the network delays and ensure zero-downtime service migration.Description
Supervisor
Xiao, YuThesis advisor
Noreikis, MariusKeywords
internet of things, fog computing, CRIU, docker containers
