Video streaming on the Internet is increasingly using Dynamic Adaptive Streaming over HTTP (DASH), which allows a client to dynamically adjust its video quality by choosing the appropriate quality level for each segment based on the current download rate. It is currently used for Video-on-Demand and live streaming in software developed by Move Networks, Apple, Netflix, Microsoft, and Adobe. However, some research has shown that DASH streaming software does not converge quickly to an appropriate video level and may be unstable or unfair in its use of bandwidth.
In this project we are developing software and algorithms to improve the performance of streaming video over the Internet. We have developed code for OMNET++ that enables the simulator to interact with the decision engine of a streaming video player. Separately, we have used a combination of analysis, dynamic programming and simulation to investigate impact of Scalable Video Coding (SVC) on the client’s quality selection policy.
This project is supported by a gift from Move Networks.