NOSSDAV 2K

Paper #32

Providing Smoother Quality Layered Video Stream

Srihari Nelakuditi, Raja R. Harinath
Ewa Kusmierek & Zhi-Li Zhang
University of Minnesota, Minneapolis

Abstract:

In recent years, one of the most popular Internet applications is web-based audio and video playback, where stored video is streamed from the server to a client on-demand. Rigid playback deadlines coupled with constraints on resources such as network bandwidth and client buffer make video delivery a challenging task [2]. These resources could be limited in such a way that it may not be possible to deliver full-quality video. In such a situation, it is desirable to minimize the degradation in the video quality while operating within the resource constraints [9]. Layered encoding is proposed to provide finer control on video quality: the video signal is split into layers and a subset of these layers is chosen such that the resource constraints are met [5]. However it is not a trivial task to select layers such that better but consistent quality playback is ensured when the network conditions are constantly varying.

In our work, we address this layer selection problem in layered video delivery and show how smoother 1 quality video play-back can be provided by utilizing the client buffer for prefetching. We first define smoothness criteria, design metrics for measuring it, and then develop off-line algorithms to maximize smoothness for the case where the network bandwidth is varying but known a priori. We then describe an adaptive algorithm for providing smoothed layered video delivery that doesnUt assume any knowledge about future bandwidth availability. The results of our experiments for measuring and comparing the performance these schemes are then presented. We conclude the paper with a brief discussion on our future work.

Line
Download a copy of this paper in either: Compressed PostScript (110K) - or - PDF (265K) format.
Line

Back to the NOSSDAV 2000 On-Line Proceedings.

Line

- Last revised: Weds Jun 27 20:56:29 EDT 2000

Line