Authors: Hongli Luo MeiLing Shyu ShuChing Chen
Publish Date: 2008/01/25
Volume: 40, Issue: 1, Pages: 111-134
Abstract
In this paper an adaptive framework for video streaming over the Internet is presented The framework is a joint design of packet scheduling and rate control with optimal bandwidth resource allocation The transmission rate is dynamically adjusted to obtain maximal utilization of the client buffer and minimal allocation of the bandwidth Under the constraint of the transmission rate a prioritized packet scheduling is designed to provide a better visual quality of video frames The packet scheduling is a refined bandwidth allocation which takes into account of varying importance of the different packets in a compressed video stream Moreover the proposed approach is scalable with increasing multimedia flows in the distributed Internet environment Comparisons are made with the most current streaming approaches to evaluate the performance of the framework using the H264 video codec The extensive simulation results show that the average Peak Signal to Noise Ratio PSNR increases in our proposed approach It provides a better quality of the decoded frames and the quality of the decoded frames changes more smoothly The achieved video quality among different users also has a lower fluctuation which indicates a fair sharing of network resources
Keywords: