2nd International IEEE/Create-Net Conference on Testbeds and Research Infrastructures for the Development of Networks and Communities

Research Article

Frame-based modeling of H.264 constrained videoconference traffic over an IP commercial platform

Cite
BibTeX Plain Text
  • @INPROCEEDINGS{10.1109/TRIDNT.2006.1649150,
    author={S.  Domoxoudis and S.  Kouremenos and A.  Drigas and V.  Loumos},
    title={Frame-based modeling of H.264 constrained videoconference traffic over an IP commercial platform},
    proceedings={2nd International IEEE/Create-Net Conference on Testbeds and Research Infrastructures for the Development of Networks and Communities},
    publisher={IEEE},
    proceedings_a={TRIDENTCOM},
    year={2006},
    month={7},
    keywords={},
    doi={10.1109/TRIDNT.2006.1649150}
}
  • S. Domoxoudis
    S. Kouremenos
    A. Drigas
    V. Loumos
    Year: 2006
    Frame-based modeling of H.264 constrained videoconference traffic over an IP commercial platform
    TRIDENTCOM
    IEEE
    DOI: 10.1109/TRIDNT.2006.1649150
S. Domoxoudis1,2,*, S. Kouremenos1,*, A. Drigas1,*, V. Loumos2,*
  • 1: National Center for Scientific Research ‘‘DEMOKRITOS’’, DTE/YE Applied, Technologies Department, P.O. Box 15310 Gr. Ag. Paraskevi, Attiki, Greece
  • 2: National Technical University of Athens ‘‘NTUA’’, School of Electrical and Computer Engineering, Multimedia Technology Laboratory, P.O. Box 15780 Gr. Zographou, Attiki, Greece
*Contact email: sdomo@imm.demokritos.gr, skourem@imm.demokritos.gr, dr@imm.demokritos.gr, loumos@cs.ntua.gr

Abstract

In this manuscript, we contribute measurement and modeling results for H.264 encoded videoconference traffic. The analysis of the frame-size autocorrelation function of extensive video data that were gathered during realistic experiments with two popular videoconference terminals (currently the only supporting the H.264 encoder) suggested that H.264 videoconference traffic is fully uncorrelated as a consequence of the constrained (in the loop) operation mode of the encoders. By using this fact and by statistically analyzing the sampled frame-size histogram, we conclude that H.264 traffic can be reasonably represented by a D/G/infin queue with deterministic arrivals according to the video frame rate and service times that can be fitted by a histogram-based model proposed on a previous study.

Published
2006-07-05
Publisher
IEEE
http://dx.doi.org/10.1109/TRIDNT.2006.1649150
Copyright © 2006–2024 IEEE