Imperial College London
Alternate

ProfessorKinLeung

Faculty of Engineering, Department of Electrical and Electronic Engineering

Tanaka Chair in Internet Technology
 
 
 
//

Contact

 

+44 (0)20 7594 6238kin.leung Website

 
 
//

Assistant

 

Miss Vanessa Rodriguez-Gonzalez +44 (0)20 7594 6267

 
//

Location

 

810aElectrical EngineeringSouth Kensington Campus

//

Summary

 

Publications

Citation

BibTex format

@article{Wang:2017:10.1109/ACCESS.2017.2665971,
author = {Wang, S and Zafer, M and Leung, KK},
doi = {10.1109/ACCESS.2017.2665971},
journal = {IEEE Access},
pages = {2514--2533},
title = {Online placement of multi-component applications in edge computing environments},
url = {http://dx.doi.org/10.1109/ACCESS.2017.2665971},
volume = {5},
year = {2017}
}
Download

RIS format (EndNote, RefMan)

TY  - JOUR
AB  - Mobile edge computing is a new cloud computingparadigm which makes use of small-sized edge-clouds to providereal-time services to users. These mobile edge-clouds (MECs)are located in close proximity to users, thus enabling users toseamlessly access applications running on MECs. Due to the coexistenceof the core (centralized) cloud, users, and one or multiplelayers of MECs, an important problem is to decide where (onwhich computational entity) to place different components of anapplication. This problem, known as the application or workloadplacement problem, is notoriously hard, and therefore, heuristicalgorithms without performance guarantees are generallyemployed in common practice, which may unknowingly sufferfrom poor performance as compared to the optimal solution.In this paper, we address the application placement problemand focus on developing algorithms with provable performancebounds. We model the user application as an application graphand the physical computing system as a physical graph, withresource demands/availabilities annotated on these graphs. Wefirst consider the placement of a linear application graph andpropose an algorithm for finding its optimal solution. Using thisresult, we then generalize the formulation and obtain onlineapproximation algorithms with polynomial-logarithmic (poly-log)competitive ratio for tree application graph placement.We jointlyconsider node and link assignment, and incorporate multipletypes of computational resources at nodes.
AU  - Wang,S
AU  - Zafer,M
AU  - Leung,KK
DO  - 10.1109/ACCESS.2017.2665971
EP  - 2533
PY  - 2017///
SN  - 2169-3536
SP  - 2514
TI  - Online placement of multi-component applications in edge computing environments
T2  - IEEE Access
UR  - http://dx.doi.org/10.1109/ACCESS.2017.2665971
UR  - http://hdl.handle.net/10044/1/44349
VL  - 5
ER  -
Download