Modeling full-length video using Markov-modulated Gamma-based framework

Research output: Contribution to journalArticle

78 Citations (Scopus)

Abstract

All traffic models for MPEG-Iike encoded variable bit rate (VBR) video can be broadly categorized into 1) data-rate models (DRMs) and 2) frame-size models (FSMs). Almost all proposed VBR traffic models are DRMs. DRMs generate only data arrival rate, and are good for estimating average packet-loss and ATM buffer overflowing probabilities, but fail to identify such details as percentage of frames affected. FSMs generate sizes of individual MPEG frames, and are good for studying frame loss rate in addition to data loss rate. Among three previously proposed FSMs: 1) one generates frame sizes for full-length movies without preserving group-of-pictures (GOP) periodicity; 2) one generates VBR video traffic for news videos from scene content description provided to it; and 3) one generates frame sizes for full-length movies without preserving size-based video-segment transitions. In this paper, we propose two FSMs that generate frame sizes for full-length VBR videos preserving both GOP periodicity and size-based video-segment transitions. First, two-pass algorithms for analysis of full-length VBR videos are presented. After two-pass analysis, these algorithms identify size-based classes of video shots into which the GOPs are partitioned. Frames in each class produce three data sets, one each for I-, B-, and P-type frames. Each of these data sets is modeled with an axis-shifted Gamma distribution. Markov renewal processes model (size-based) video segment transitions. We have used QQ plots to show visual similarity of model-generated VBR video data sets with original data set. Leaky-bucket simulation study has been used to show similarity of data and frame loss rates between model-generated VBR videos and original video. Our study of frame-based VBR video revealed that even a low data-loss rate could affect a large fraction of I frames, causing a significant degradation of the quality of transmitted video.

Original languageEnglish (US)
Pages (from-to)638-649
Number of pages12
JournalIEEE/ACM Transactions on Networking
Volume11
Issue number4
DOIs
StatePublished - Aug 1 2003

Fingerprint

Automatic teller machines
Packet loss
Markov processes
Degradation

Keywords

  • Frame-size traffic model
  • Gamma distribution
  • Leaky-bucket simulation
  • MPEG
  • QQ plot
  • Variable bit rate (VBR) video

ASJC Scopus subject areas

  • Software
  • Computer Science Applications
  • Computer Networks and Communications
  • Electrical and Electronic Engineering

Cite this

Modeling full-length video using Markov-modulated Gamma-based framework. / Sarkar, Uttam K.; Ramakrishnan, Subramanian; Sarkar, Dilip.

In: IEEE/ACM Transactions on Networking, Vol. 11, No. 4, 01.08.2003, p. 638-649.

Research output: Contribution to journalArticle

@article{70dc6679ef0c4505811a36e355f903a8,
title = "Modeling full-length video using Markov-modulated Gamma-based framework",
abstract = "All traffic models for MPEG-Iike encoded variable bit rate (VBR) video can be broadly categorized into 1) data-rate models (DRMs) and 2) frame-size models (FSMs). Almost all proposed VBR traffic models are DRMs. DRMs generate only data arrival rate, and are good for estimating average packet-loss and ATM buffer overflowing probabilities, but fail to identify such details as percentage of frames affected. FSMs generate sizes of individual MPEG frames, and are good for studying frame loss rate in addition to data loss rate. Among three previously proposed FSMs: 1) one generates frame sizes for full-length movies without preserving group-of-pictures (GOP) periodicity; 2) one generates VBR video traffic for news videos from scene content description provided to it; and 3) one generates frame sizes for full-length movies without preserving size-based video-segment transitions. In this paper, we propose two FSMs that generate frame sizes for full-length VBR videos preserving both GOP periodicity and size-based video-segment transitions. First, two-pass algorithms for analysis of full-length VBR videos are presented. After two-pass analysis, these algorithms identify size-based classes of video shots into which the GOPs are partitioned. Frames in each class produce three data sets, one each for I-, B-, and P-type frames. Each of these data sets is modeled with an axis-shifted Gamma distribution. Markov renewal processes model (size-based) video segment transitions. We have used QQ plots to show visual similarity of model-generated VBR video data sets with original data set. Leaky-bucket simulation study has been used to show similarity of data and frame loss rates between model-generated VBR videos and original video. Our study of frame-based VBR video revealed that even a low data-loss rate could affect a large fraction of I frames, causing a significant degradation of the quality of transmitted video.",
keywords = "Frame-size traffic model, Gamma distribution, Leaky-bucket simulation, MPEG, QQ plot, Variable bit rate (VBR) video",
author = "Sarkar, {Uttam K.} and Subramanian Ramakrishnan and Dilip Sarkar",
year = "2003",
month = "8",
day = "1",
doi = "10.1109/TNET.2003.815292",
language = "English (US)",
volume = "11",
pages = "638--649",
journal = "IEEE/ACM Transactions on Networking",
issn = "1063-6692",
publisher = "Institute of Electrical and Electronics Engineers Inc.",
number = "4",

}

TY - JOUR

T1 - Modeling full-length video using Markov-modulated Gamma-based framework

AU - Sarkar, Uttam K.

AU - Ramakrishnan, Subramanian

AU - Sarkar, Dilip

PY - 2003/8/1

Y1 - 2003/8/1

N2 - All traffic models for MPEG-Iike encoded variable bit rate (VBR) video can be broadly categorized into 1) data-rate models (DRMs) and 2) frame-size models (FSMs). Almost all proposed VBR traffic models are DRMs. DRMs generate only data arrival rate, and are good for estimating average packet-loss and ATM buffer overflowing probabilities, but fail to identify such details as percentage of frames affected. FSMs generate sizes of individual MPEG frames, and are good for studying frame loss rate in addition to data loss rate. Among three previously proposed FSMs: 1) one generates frame sizes for full-length movies without preserving group-of-pictures (GOP) periodicity; 2) one generates VBR video traffic for news videos from scene content description provided to it; and 3) one generates frame sizes for full-length movies without preserving size-based video-segment transitions. In this paper, we propose two FSMs that generate frame sizes for full-length VBR videos preserving both GOP periodicity and size-based video-segment transitions. First, two-pass algorithms for analysis of full-length VBR videos are presented. After two-pass analysis, these algorithms identify size-based classes of video shots into which the GOPs are partitioned. Frames in each class produce three data sets, one each for I-, B-, and P-type frames. Each of these data sets is modeled with an axis-shifted Gamma distribution. Markov renewal processes model (size-based) video segment transitions. We have used QQ plots to show visual similarity of model-generated VBR video data sets with original data set. Leaky-bucket simulation study has been used to show similarity of data and frame loss rates between model-generated VBR videos and original video. Our study of frame-based VBR video revealed that even a low data-loss rate could affect a large fraction of I frames, causing a significant degradation of the quality of transmitted video.

AB - All traffic models for MPEG-Iike encoded variable bit rate (VBR) video can be broadly categorized into 1) data-rate models (DRMs) and 2) frame-size models (FSMs). Almost all proposed VBR traffic models are DRMs. DRMs generate only data arrival rate, and are good for estimating average packet-loss and ATM buffer overflowing probabilities, but fail to identify such details as percentage of frames affected. FSMs generate sizes of individual MPEG frames, and are good for studying frame loss rate in addition to data loss rate. Among three previously proposed FSMs: 1) one generates frame sizes for full-length movies without preserving group-of-pictures (GOP) periodicity; 2) one generates VBR video traffic for news videos from scene content description provided to it; and 3) one generates frame sizes for full-length movies without preserving size-based video-segment transitions. In this paper, we propose two FSMs that generate frame sizes for full-length VBR videos preserving both GOP periodicity and size-based video-segment transitions. First, two-pass algorithms for analysis of full-length VBR videos are presented. After two-pass analysis, these algorithms identify size-based classes of video shots into which the GOPs are partitioned. Frames in each class produce three data sets, one each for I-, B-, and P-type frames. Each of these data sets is modeled with an axis-shifted Gamma distribution. Markov renewal processes model (size-based) video segment transitions. We have used QQ plots to show visual similarity of model-generated VBR video data sets with original data set. Leaky-bucket simulation study has been used to show similarity of data and frame loss rates between model-generated VBR videos and original video. Our study of frame-based VBR video revealed that even a low data-loss rate could affect a large fraction of I frames, causing a significant degradation of the quality of transmitted video.

KW - Frame-size traffic model

KW - Gamma distribution

KW - Leaky-bucket simulation

KW - MPEG

KW - QQ plot

KW - Variable bit rate (VBR) video

UR - http://www.scopus.com/inward/record.url?scp=0042416624&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0042416624&partnerID=8YFLogxK

U2 - 10.1109/TNET.2003.815292

DO - 10.1109/TNET.2003.815292

M3 - Article

VL - 11

SP - 638

EP - 649

JO - IEEE/ACM Transactions on Networking

JF - IEEE/ACM Transactions on Networking

SN - 1063-6692

IS - 4

ER -