Fitting World-Wide Web Request Traces with the EM-Algorithm

Rachid El Abdouni Khayari, R. Sadre, Boudewijn R.H.M. Haverkort

Research output: Contribution to journalArticleScientificpeer-review

Abstract

In recent years, various researchers have shown that network traffic that is due to world-wide web transfers shows characteristics of self-similarity and it has been argued that this can be explained by the heavy-tailedness of many of the involved distributions. Considering these facts, developing methods that are able to handle self-similarity and heavy-tailedness is of great importance for network capacity planning purposes. However, heavy-tailed distributions cannot be used so easily for analytical or numerical evaluation studies. To overcome this problem, in this paper, we approximate the empirical distributions by analytically more tractable, that is, hyper-exponential distributions. For that purpose, we present a new fitting algorithm based on the expectation-maximisation and show it to perform well both for pure traffic statistics as well as in queuing studies.
Original languageEnglish
Pages (from-to)175-191
Number of pages17
JournalPerformance Evaluation
Volume52
Issue number2-3
DOIs
Publication statusPublished - Apr 2003
Externally publishedYes

Fingerprint

Self-similarity
EM Algorithm
World Wide Web
Trace
Statistics
Network Planning
Planning
Capacity Planning
Heavy-tailed Distribution
Expectation Maximization
Queuing
Empirical Distribution
Network Traffic
Exponential distribution
Traffic
Evaluation

Keywords

  • Traffic characterisation
  • Hyper-exponential distributions
  • IR-61888
  • EWI-10922
  • EM-fitting
  • Queuing analysis
  • World Wide Web
  • ML-fitting
  • Heavy-tailed distributions

Cite this

El Abdouni Khayari, Rachid ; Sadre, R. ; Haverkort, Boudewijn R.H.M. / Fitting World-Wide Web Request Traces with the EM-Algorithm. In: Performance Evaluation. 2003 ; Vol. 52, No. 2-3. pp. 175-191.
@article{928e4a155a614edeb109d897c33bfb8d,
title = "Fitting World-Wide Web Request Traces with the EM-Algorithm",
abstract = "In recent years, various researchers have shown that network traffic that is due to world-wide web transfers shows characteristics of self-similarity and it has been argued that this can be explained by the heavy-tailedness of many of the involved distributions. Considering these facts, developing methods that are able to handle self-similarity and heavy-tailedness is of great importance for network capacity planning purposes. However, heavy-tailed distributions cannot be used so easily for analytical or numerical evaluation studies. To overcome this problem, in this paper, we approximate the empirical distributions by analytically more tractable, that is, hyper-exponential distributions. For that purpose, we present a new fitting algorithm based on the expectation-maximisation and show it to perform well both for pure traffic statistics as well as in queuing studies.",
keywords = "Traffic characterisation, Hyper-exponential distributions, IR-61888, EWI-10922, EM-fitting, Queuing analysis, World Wide Web, ML-fitting, Heavy-tailed distributions",
author = "{El Abdouni Khayari}, Rachid and R. Sadre and Haverkort, {Boudewijn R.H.M.}",
year = "2003",
month = "4",
doi = "10.1016/S0166-5316(02)00179-7",
language = "English",
volume = "52",
pages = "175--191",
journal = "Performance Evaluation",
issn = "0166-5316",
publisher = "Elsevier",
number = "2-3",

}

Fitting World-Wide Web Request Traces with the EM-Algorithm. / El Abdouni Khayari, Rachid; Sadre, R.; Haverkort, Boudewijn R.H.M.

In: Performance Evaluation, Vol. 52, No. 2-3, 04.2003, p. 175-191.

Research output: Contribution to journalArticleScientificpeer-review

TY - JOUR

T1 - Fitting World-Wide Web Request Traces with the EM-Algorithm

AU - El Abdouni Khayari, Rachid

AU - Sadre, R.

AU - Haverkort, Boudewijn R.H.M.

PY - 2003/4

Y1 - 2003/4

N2 - In recent years, various researchers have shown that network traffic that is due to world-wide web transfers shows characteristics of self-similarity and it has been argued that this can be explained by the heavy-tailedness of many of the involved distributions. Considering these facts, developing methods that are able to handle self-similarity and heavy-tailedness is of great importance for network capacity planning purposes. However, heavy-tailed distributions cannot be used so easily for analytical or numerical evaluation studies. To overcome this problem, in this paper, we approximate the empirical distributions by analytically more tractable, that is, hyper-exponential distributions. For that purpose, we present a new fitting algorithm based on the expectation-maximisation and show it to perform well both for pure traffic statistics as well as in queuing studies.

AB - In recent years, various researchers have shown that network traffic that is due to world-wide web transfers shows characteristics of self-similarity and it has been argued that this can be explained by the heavy-tailedness of many of the involved distributions. Considering these facts, developing methods that are able to handle self-similarity and heavy-tailedness is of great importance for network capacity planning purposes. However, heavy-tailed distributions cannot be used so easily for analytical or numerical evaluation studies. To overcome this problem, in this paper, we approximate the empirical distributions by analytically more tractable, that is, hyper-exponential distributions. For that purpose, we present a new fitting algorithm based on the expectation-maximisation and show it to perform well both for pure traffic statistics as well as in queuing studies.

KW - Traffic characterisation

KW - Hyper-exponential distributions

KW - IR-61888

KW - EWI-10922

KW - EM-fitting

KW - Queuing analysis

KW - World Wide Web

KW - ML-fitting

KW - Heavy-tailed distributions

U2 - 10.1016/S0166-5316(02)00179-7

DO - 10.1016/S0166-5316(02)00179-7

M3 - Article

VL - 52

SP - 175

EP - 191

JO - Performance Evaluation

JF - Performance Evaluation

SN - 0166-5316

IS - 2-3

ER -