Traffic modeling and QoS prediction for MPEG-coded video services over ATM networks using scene level statistical characteristics

  • Authors:
  • Sang-Jo Yoo;Seong-Dae Kim

  • Affiliations:
  • -;-

  • Venue:
  • Journal of High Speed Networks
  • Year:
  • 2000

Quantified Score

Hi-index 0.00

Visualization

Abstract

In this paper, we propose a new traffic model and its queueingsolution for MPEG-coded video sequences, which are now widely usedfor many types of video applications. The MPEG variable bit rate(VBR) video traffic has both very strong short-range correlationsand long-range dependence. To capture the long-range dependence, wemodel the overall scene-level cell rate flow with a scene-levelstate transition diagram. At the same time, we use a frame-levelchain to reflect the short-range correlation effects between theframes of the same frame type and between the different frame typeswithin a scene state. Thus scene changes and frame sequences aremodeled by a two-dimensional continuous time Markov chain. Weprovide an analytic queueing solution in terms of the maximum celltransfer delay and the cell loss ratio. The simulation results showthat our analytical model accurately approximates the queueingperformance of the original videos. Finally, with an asymptoticapproximation, we provide a method for estimating the effectivebandwidth for MPEG video sources with a guarantee of the usersquality of service requirements.