Approximate analysis of a dynamic scheduler for self-similar video traffic in ATM networks

T Kocak

Research output: Contribution to journalArticle (Academic Journal)peer-review

2 Citations (Scopus)

Abstract

Performance evaluation methods that have been used so far failed to capture the real characteristics of network traffic. This is especially true when the service times are general such as the variable bit rate (VBR) video, which is proven to be self-similar. In the paper, diffusion approximation methods are proposed to analyse a dynamic scheduler for self-similar VBR video traffic in asynchronous transfer mode (ATM) networks. The proposed scheme employs a coupled queueing system that has been used extensively in the modelling of computer and communication systems. Diffusion approximations methods are used to decouple a queueing system which represents an ATM network node into separate G/G/1 queues. Real MPEG video traces are used in the discrete event simulation. Results are compared with the approximation, and are found to work very well under different traffic conditions.
Translated title of the contributionApproximate analysis of a dynamic scheduler for self-similar video traffic in ATM networks
Original languageEnglish
Pages (from-to)189 - 194
Number of pages6
JournalIEE Proceedings - Communications
Volume153
Issue number2
DOIs
Publication statusPublished - Apr 2006

Bibliographical note

Publisher: Institution of Electrical Engineers (IEE)
Rose publication type: Journal article

Keywords

  • approximation theory
  • asynchronous transfer mode

Fingerprint Dive into the research topics of 'Approximate analysis of a dynamic scheduler for self-similar video traffic in ATM networks'. Together they form a unique fingerprint.

Cite this