On the characterization of VBR MPEG streams

Marwan M Krunz, Satish K. Tripathi

Research output: Chapter in Book/Report/Conference proceedingChapter

98 Citations (Scopus)

Abstract

We present a comprehensive model for variable-bit-rate MPEG video streams. This model captures the bit-rate variations at multiple time scales. Long-term variations are captured by incorporating scene changes, which are most noticeable in the fluctuations of I frames. The size of an I frame is modeled by the sum of two random components: a scene-related component and an AR(2) component that accounts for the fluctuations within a scene. Two random processes of i.i.d. rvs are used to model the sizes of P and B frames, respectively. The complete model is then obtained by intermixing the three sub-models according to a given GOP pattern. It is shown that the composite model exhibits long-range dependence (LRD) in the sense that its autocorrelation function is non-summable. The LRD behavior is caused by the repetitive GOP pattern which induces periodic cross-correlations between different types of frames. Using standard statistical methods, we successfully fit our model to several empirical video traces. We then study the queueing performance for video traffic at a statistical multiplexer. The results show that the model is sufficiently accurate in predicting the queueing performance for real video streams.

Original languageEnglish (US)
Title of host publicationPerformance Evaluation Review
Editors Anon
PublisherACM
Pages192-202
Number of pages11
Volume25
Edition1
StatePublished - Jun 1997
EventProceedings of the 1997 ACM Sigmetrics International Conference on Measurement and Modeling of Computer Systems - Seattle, WA, USA
Duration: Jun 15 1997Jun 18 1997

Other

OtherProceedings of the 1997 ACM Sigmetrics International Conference on Measurement and Modeling of Computer Systems
CitySeattle, WA, USA
Period6/15/976/18/97

Fingerprint

Random processes
Autocorrelation
Statistical methods
Composite materials

ASJC Scopus subject areas

  • Hardware and Architecture

Cite this

Krunz, M. M., & Tripathi, S. K. (1997). On the characterization of VBR MPEG streams. In Anon (Ed.), Performance Evaluation Review (1 ed., Vol. 25, pp. 192-202). ACM.

On the characterization of VBR MPEG streams. / Krunz, Marwan M; Tripathi, Satish K.

Performance Evaluation Review. ed. / Anon. Vol. 25 1. ed. ACM, 1997. p. 192-202.

Research output: Chapter in Book/Report/Conference proceedingChapter

Krunz, MM & Tripathi, SK 1997, On the characterization of VBR MPEG streams. in Anon (ed.), Performance Evaluation Review. 1 edn, vol. 25, ACM, pp. 192-202, Proceedings of the 1997 ACM Sigmetrics International Conference on Measurement and Modeling of Computer Systems, Seattle, WA, USA, 6/15/97.
Krunz MM, Tripathi SK. On the characterization of VBR MPEG streams. In Anon, editor, Performance Evaluation Review. 1 ed. Vol. 25. ACM. 1997. p. 192-202
Krunz, Marwan M ; Tripathi, Satish K. / On the characterization of VBR MPEG streams. Performance Evaluation Review. editor / Anon. Vol. 25 1. ed. ACM, 1997. pp. 192-202
@inbook{5b5afed818614acd9f72365f6aea13dd,
title = "On the characterization of VBR MPEG streams",
abstract = "We present a comprehensive model for variable-bit-rate MPEG video streams. This model captures the bit-rate variations at multiple time scales. Long-term variations are captured by incorporating scene changes, which are most noticeable in the fluctuations of I frames. The size of an I frame is modeled by the sum of two random components: a scene-related component and an AR(2) component that accounts for the fluctuations within a scene. Two random processes of i.i.d. rvs are used to model the sizes of P and B frames, respectively. The complete model is then obtained by intermixing the three sub-models according to a given GOP pattern. It is shown that the composite model exhibits long-range dependence (LRD) in the sense that its autocorrelation function is non-summable. The LRD behavior is caused by the repetitive GOP pattern which induces periodic cross-correlations between different types of frames. Using standard statistical methods, we successfully fit our model to several empirical video traces. We then study the queueing performance for video traffic at a statistical multiplexer. The results show that the model is sufficiently accurate in predicting the queueing performance for real video streams.",
author = "Krunz, {Marwan M} and Tripathi, {Satish K.}",
year = "1997",
month = "6",
language = "English (US)",
volume = "25",
pages = "192--202",
editor = "Anon",
booktitle = "Performance Evaluation Review",
publisher = "ACM",
edition = "1",

}

TY - CHAP

T1 - On the characterization of VBR MPEG streams

AU - Krunz, Marwan M

AU - Tripathi, Satish K.

PY - 1997/6

Y1 - 1997/6

N2 - We present a comprehensive model for variable-bit-rate MPEG video streams. This model captures the bit-rate variations at multiple time scales. Long-term variations are captured by incorporating scene changes, which are most noticeable in the fluctuations of I frames. The size of an I frame is modeled by the sum of two random components: a scene-related component and an AR(2) component that accounts for the fluctuations within a scene. Two random processes of i.i.d. rvs are used to model the sizes of P and B frames, respectively. The complete model is then obtained by intermixing the three sub-models according to a given GOP pattern. It is shown that the composite model exhibits long-range dependence (LRD) in the sense that its autocorrelation function is non-summable. The LRD behavior is caused by the repetitive GOP pattern which induces periodic cross-correlations between different types of frames. Using standard statistical methods, we successfully fit our model to several empirical video traces. We then study the queueing performance for video traffic at a statistical multiplexer. The results show that the model is sufficiently accurate in predicting the queueing performance for real video streams.

AB - We present a comprehensive model for variable-bit-rate MPEG video streams. This model captures the bit-rate variations at multiple time scales. Long-term variations are captured by incorporating scene changes, which are most noticeable in the fluctuations of I frames. The size of an I frame is modeled by the sum of two random components: a scene-related component and an AR(2) component that accounts for the fluctuations within a scene. Two random processes of i.i.d. rvs are used to model the sizes of P and B frames, respectively. The complete model is then obtained by intermixing the three sub-models according to a given GOP pattern. It is shown that the composite model exhibits long-range dependence (LRD) in the sense that its autocorrelation function is non-summable. The LRD behavior is caused by the repetitive GOP pattern which induces periodic cross-correlations between different types of frames. Using standard statistical methods, we successfully fit our model to several empirical video traces. We then study the queueing performance for video traffic at a statistical multiplexer. The results show that the model is sufficiently accurate in predicting the queueing performance for real video streams.

UR - http://www.scopus.com/inward/record.url?scp=0031169609&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0031169609&partnerID=8YFLogxK

M3 - Chapter

AN - SCOPUS:0031169609

VL - 25

SP - 192

EP - 202

BT - Performance Evaluation Review

A2 - Anon, null

PB - ACM

ER -