Cover Image

Network traffic engineering using linear regression approach / Vathsala Devi Kunalan

Quality of Service (QoS) routing is becoming a very important criteria in Internet for supporting the ever-increasing diverse multimedia applications that demand very high level quality of service from the underlying network. To achieve this, QoS-enabled routers need to maintain accurate view of the...

Full description

Saved in:
Bibliographic Details
Main Author: Vathsala Devi, Kunalan
Format: Thesis
Published: 2006
Subjects:
QA75 Electronic computers. Computer science
Online Access:http://studentsrepo.um.edu.my/10773/1/Vathsala_Devi_Kunalan_%E2%80%93_Dissertation.pdf
http://studentsrepo.um.edu.my/10773/
Tags: Add Tag
No Tags, Be the first to tag this record!
id my.um.stud.10773
record_format eprints
spelling my.um.stud.107732020-08-17T00:14:27Z Network traffic engineering using linear regression approach / Vathsala Devi Kunalan Vathsala Devi, Kunalan QA75 Electronic computers. Computer science Quality of Service (QoS) routing is becoming a very important criteria in Internet for supporting the ever-increasing diverse multimedia applications that demand very high level quality of service from the underlying network. To achieve this, QoS-enabled routers need to maintain accurate view of the network resource availability by exchanging state information among themselves at appropriate intervals. Frequent dissemination of global network state information introduces two major problems: increased computational costs and higher protocol overheads. In this thesis, a new mechanism that disseminates link state updates based on the bandwidth utilization trend of a link, called TE-LR (Traffic Engineering using Linear Regression), is proposed. The main idea of the mechanism is to sample the bandwidth utilization ratio of a link at regular intervals and use these sampled data to construct linear regression line equations. The tangent value obtained from the equation is used to decide when to send link state updates. Simulation results showed that the proposed mechanism had been successful in reducing the update message overheads by almost 78% with minimal impact to other parameters such as packet loss ratio, , average link utilization and average end-to-end delay. 2006 Thesis NonPeerReviewed application/pdf http://studentsrepo.um.edu.my/10773/1/Vathsala_Devi_Kunalan_%E2%80%93_Dissertation.pdf Vathsala Devi, Kunalan (2006) Network traffic engineering using linear regression approach / Vathsala Devi Kunalan. Masters thesis, University of Malaya. http://studentsrepo.um.edu.my/10773/
institution Universiti Malaya
building UM Library
collection Institutional Repository
continent Asia
country Malaysia
content_provider Universiti Malaya
content_source UM Student Repository
url_provider http://studentsrepo.um.edu.my/
topic QA75 Electronic computers. Computer science
spellingShingle QA75 Electronic computers. Computer science
Vathsala Devi, Kunalan
Network traffic engineering using linear regression approach / Vathsala Devi Kunalan
description Quality of Service (QoS) routing is becoming a very important criteria in Internet for supporting the ever-increasing diverse multimedia applications that demand very high level quality of service from the underlying network. To achieve this, QoS-enabled routers need to maintain accurate view of the network resource availability by exchanging state information among themselves at appropriate intervals. Frequent dissemination of global network state information introduces two major problems: increased computational costs and higher protocol overheads. In this thesis, a new mechanism that disseminates link state updates based on the bandwidth utilization trend of a link, called TE-LR (Traffic Engineering using Linear Regression), is proposed. The main idea of the mechanism is to sample the bandwidth utilization ratio of a link at regular intervals and use these sampled data to construct linear regression line equations. The tangent value obtained from the equation is used to decide when to send link state updates. Simulation results showed that the proposed mechanism had been successful in reducing the update message overheads by almost 78% with minimal impact to other parameters such as packet loss ratio, , average link utilization and average end-to-end delay.
format Thesis
author Vathsala Devi, Kunalan
author_facet Vathsala Devi, Kunalan
author_sort Vathsala Devi, Kunalan
title Network traffic engineering using linear regression approach / Vathsala Devi Kunalan
title_short Network traffic engineering using linear regression approach / Vathsala Devi Kunalan
title_full Network traffic engineering using linear regression approach / Vathsala Devi Kunalan
title_fullStr Network traffic engineering using linear regression approach / Vathsala Devi Kunalan
title_full_unstemmed Network traffic engineering using linear regression approach / Vathsala Devi Kunalan
title_sort network traffic engineering using linear regression approach / vathsala devi kunalan
publishDate 2006
url http://studentsrepo.um.edu.my/10773/1/Vathsala_Devi_Kunalan_%E2%80%93_Dissertation.pdf
http://studentsrepo.um.edu.my/10773/
_version_ 1738506408901476352
score 13.211869

Similar Items