On the propagation of long-range dependence in the Internet

A. Veres, Zs Kenesi, S. Molnár, G. Vattay

Research output: Chapter in Book/Report/Conference proceedingChapter

24 Citations (Scopus)

Abstract

This paper analyzes how TCP congestion control can propagate self-similarity between distant areas of the Internet. This property of TCP is due to its congestion control algorithm, which adapts to self-similar fluctuations on several timescales. The mechanisms and limitations of this propagation are investigated, and it is demonstrated that if a TCP connection shares a bottleneck link with a self-similar background traffic flow, it propagates the correlation structure of the background traffic flow above a characteristic timescale. The cut-off timescale depends on the end-to-end path properties, e.g., round-trip time and average window size. It is also demonstrated that even short TCP connections can propagate long-range correlations effectively. Our analysis reveals that if congestion periods in a connection's hops are long-range dependent, then the end-user perceived end-to-end traffic is also long-range dependent and it is characterized by the largest Hurst exponent. Furthermore, it is shown that self-similarity of one TCP stream can be passed on to other TCP streams that it is multiplexed with. These mechanisms complement the widespread scaling phenomena reported in a number of recent papers. Our arguments are supported with a combination of analytic techniques, simulations and statistical analyses of real Internet traffic measurements.

Original languageEnglish
Title of host publicationComputer Communication Review
PublisherACM
Pages243-254
Number of pages12
Volume30
Edition4
Publication statusPublished - 2000
EventProceedings of ACM SIGCOMM 2000 Conference - Stockholm, Swed
Duration: Aug 28 2000Sep 1 2000

Other

OtherProceedings of ACM SIGCOMM 2000 Conference
CityStockholm, Swed
Period8/28/009/1/00

Fingerprint

Internet
Congestion control (communication)

ASJC Scopus subject areas

  • Information Systems

Cite this

Veres, A., Kenesi, Z., Molnár, S., & Vattay, G. (2000). On the propagation of long-range dependence in the Internet. In Computer Communication Review (4 ed., Vol. 30, pp. 243-254). ACM.

On the propagation of long-range dependence in the Internet. / Veres, A.; Kenesi, Zs; Molnár, S.; Vattay, G.

Computer Communication Review. Vol. 30 4. ed. ACM, 2000. p. 243-254.

Research output: Chapter in Book/Report/Conference proceedingChapter

Veres, A, Kenesi, Z, Molnár, S & Vattay, G 2000, On the propagation of long-range dependence in the Internet. in Computer Communication Review. 4 edn, vol. 30, ACM, pp. 243-254, Proceedings of ACM SIGCOMM 2000 Conference, Stockholm, Swed, 8/28/00.
Veres A, Kenesi Z, Molnár S, Vattay G. On the propagation of long-range dependence in the Internet. In Computer Communication Review. 4 ed. Vol. 30. ACM. 2000. p. 243-254
Veres, A. ; Kenesi, Zs ; Molnár, S. ; Vattay, G. / On the propagation of long-range dependence in the Internet. Computer Communication Review. Vol. 30 4. ed. ACM, 2000. pp. 243-254
@inbook{d144d61b35d546f79685ab5c8d1bba7b,
title = "On the propagation of long-range dependence in the Internet",
abstract = "This paper analyzes how TCP congestion control can propagate self-similarity between distant areas of the Internet. This property of TCP is due to its congestion control algorithm, which adapts to self-similar fluctuations on several timescales. The mechanisms and limitations of this propagation are investigated, and it is demonstrated that if a TCP connection shares a bottleneck link with a self-similar background traffic flow, it propagates the correlation structure of the background traffic flow above a characteristic timescale. The cut-off timescale depends on the end-to-end path properties, e.g., round-trip time and average window size. It is also demonstrated that even short TCP connections can propagate long-range correlations effectively. Our analysis reveals that if congestion periods in a connection's hops are long-range dependent, then the end-user perceived end-to-end traffic is also long-range dependent and it is characterized by the largest Hurst exponent. Furthermore, it is shown that self-similarity of one TCP stream can be passed on to other TCP streams that it is multiplexed with. These mechanisms complement the widespread scaling phenomena reported in a number of recent papers. Our arguments are supported with a combination of analytic techniques, simulations and statistical analyses of real Internet traffic measurements.",
author = "A. Veres and Zs Kenesi and S. Moln{\'a}r and G. Vattay",
year = "2000",
language = "English",
volume = "30",
pages = "243--254",
booktitle = "Computer Communication Review",
publisher = "ACM",
edition = "4",

}

TY - CHAP

T1 - On the propagation of long-range dependence in the Internet

AU - Veres, A.

AU - Kenesi, Zs

AU - Molnár, S.

AU - Vattay, G.

PY - 2000

Y1 - 2000

N2 - This paper analyzes how TCP congestion control can propagate self-similarity between distant areas of the Internet. This property of TCP is due to its congestion control algorithm, which adapts to self-similar fluctuations on several timescales. The mechanisms and limitations of this propagation are investigated, and it is demonstrated that if a TCP connection shares a bottleneck link with a self-similar background traffic flow, it propagates the correlation structure of the background traffic flow above a characteristic timescale. The cut-off timescale depends on the end-to-end path properties, e.g., round-trip time and average window size. It is also demonstrated that even short TCP connections can propagate long-range correlations effectively. Our analysis reveals that if congestion periods in a connection's hops are long-range dependent, then the end-user perceived end-to-end traffic is also long-range dependent and it is characterized by the largest Hurst exponent. Furthermore, it is shown that self-similarity of one TCP stream can be passed on to other TCP streams that it is multiplexed with. These mechanisms complement the widespread scaling phenomena reported in a number of recent papers. Our arguments are supported with a combination of analytic techniques, simulations and statistical analyses of real Internet traffic measurements.

AB - This paper analyzes how TCP congestion control can propagate self-similarity between distant areas of the Internet. This property of TCP is due to its congestion control algorithm, which adapts to self-similar fluctuations on several timescales. The mechanisms and limitations of this propagation are investigated, and it is demonstrated that if a TCP connection shares a bottleneck link with a self-similar background traffic flow, it propagates the correlation structure of the background traffic flow above a characteristic timescale. The cut-off timescale depends on the end-to-end path properties, e.g., round-trip time and average window size. It is also demonstrated that even short TCP connections can propagate long-range correlations effectively. Our analysis reveals that if congestion periods in a connection's hops are long-range dependent, then the end-user perceived end-to-end traffic is also long-range dependent and it is characterized by the largest Hurst exponent. Furthermore, it is shown that self-similarity of one TCP stream can be passed on to other TCP streams that it is multiplexed with. These mechanisms complement the widespread scaling phenomena reported in a number of recent papers. Our arguments are supported with a combination of analytic techniques, simulations and statistical analyses of real Internet traffic measurements.

UR - http://www.scopus.com/inward/record.url?scp=0033656192&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0033656192&partnerID=8YFLogxK

M3 - Chapter

AN - SCOPUS:0033656192

VL - 30

SP - 243

EP - 254

BT - Computer Communication Review

PB - ACM

ER -