Mario Baldi |
|
Applications that require real-time interaction among their users are gaining importance and diffusion as computer networks become increasingly powerful and ubiquitous. Many such applications impose very stringent requirements on the network; among the applications today widely deployed, videoconferencing is most demanding.
In order for the participants in a videoconference call to interact naturally, the end-to-end delay should be below human perception; even though an objective and unique figure cannot be set, 100 ms is widely recognized as the desired one way delay requirement for interaction. Since the global propagation delay alone can be about 100 ms, the actual end-to-end delay budget available to the system designer (excluding propagation delay) is very limited (e.g., no more than 10 ms).
We identify the components of the end-to-end delay in various configurations with the objective of understanding how it can be kept below the desired 100 ms bound.
After an overview of the distinctive features of interactive multimedia applications, their requirements on the network are discussed and limited delay and delay variation (Jitter) identified as the most demanding requirements. The most common techniques for controlling service quality are presented and the two major frameworks for their deployment within IP networks, i.e., Integrated Services (IntServ) and Differentiated Services (DiffServ), outlined. The tutorial also studies the implications of the architecture of packet switches on delay bounds.
Then the tutorial delves into more detail by analyzing the end-to-end delay components of a videoconferencing system step-by-step, through six system configurations obtained by combining three generic network architectures with two video encoding schemes: the transmission of raw video and variable bit rate (VBR) MPEG video encoding over (i) circuit switching, (ii) synchronous packet switching, and (iii) asynchronous packet switching.
Various queuing and scheduling algorithms for asynchronous and synchronous packet networks will be analyzed and compared.
The tutorial finally shows that having a global common time reference, together with an innovative queuing and scheduling scheme called time-driven priority (TDP) and VBR MPEG video encoding, provides adequate end-to-end delay, which is
The participants are expected to have basic knowledge on packet switching and the Internet Protocol Suite. Depending on the background of the audience on multimedia, the first part of the tutorial can be organized as a quick overview of the general context to which the remainder, or as an in-depth discussion.
M. Baldi, F. Risso, "Efficiency of Packet Voice with Deterministic Delay," IEEE Communications Magazine, Vol. 38, No. 5, May 2000, pp. 170-177. (earlier version in gzipped postscript and PDF format)
Some of the documents linked to this page are readable using Acrobat PDF Reader that you can download by clicking the icon aside |
Last Update in August 2006