How good (or bad) is shortest path routing in layered networks

Ronald G. Addie, David Fatseas, Yu Peng, Fan Li, Moshe Zukerman

Research output: Chapters, Conference Papers, Creative and Literary WorksRGC 32 - Refereed conference paper (with host publication)peer-review

1 Citation (Scopus)

Abstract

Shortest path routing is used throughout the Internet. If link costs accurately reflect real costs of installation and maintenance, and the total network cost is accurately modelled as the sum of link costs, shortest path routing can be expected to result in near optimal cost. In this paper, we consider layered networks which use shortest path routing in all layers. We study the sensitivity of shortest path routing in all the layers to link cost errors in terms of the effect on the total network cost. Examples of multilayer shortest path implementation on realistic networks show that the optimal solution can be very sensitive to link cost errors, particularly when switching cost is a significant part of total network cost, as is expected to be the case, more and more, in the future. This signifies the importance of considering accurate link costs in network design which contradicts common views that good network design based on shortest path is possible despite the presence of significant link cost errors. © 2012 IEEE.
Original languageEnglish
Title of host publicationAustralasian Telecommunication Networks and Applications Conference, ATNAC 2012
DOIs
Publication statusPublished - 2012
Event2012 Australasian Telecommunication Networks and Applications Conference, ATNAC 2012 - Brisbane, QLD, Australia
Duration: 7 Nov 20129 Nov 2012

Conference

Conference2012 Australasian Telecommunication Networks and Applications Conference, ATNAC 2012
Country/TerritoryAustralia
CityBrisbane, QLD
Period7/11/129/11/12

Fingerprint

Dive into the research topics of 'How good (or bad) is shortest path routing in layered networks'. Together they form a unique fingerprint.

Cite this