Grants and Contracts Details
Description
Project Summary
The content and service provided by the web have penetrated to every discipline of scientific
research. Making information available to a rapidly growing user population with a high service
quality is quickly becoming a very important and challenging problem. More recently, distribution
of video and audio over the Internet has attracted more attentions. Multimedia streaming brings
new challenges to content distribution. The system usually works fine with a small number of
users, but quickly becomes inaccessible when the number of users increases. In this project, we
propose to use overlay multicast (also known as application-layer multicast) as the basic structure
for the streaming task. Rather than letting all users get the multimedia content from the origin
server, we establish a delivery tree among the application end points. This will reduce the load
on the origin server and can make the streaming content available to a large population. One key
problem that has not been solved is the failures or unexpected departures of application end points.
Once nonleaf application end points leave the delivery tree, all downstream nodes will be affected.
We will investigate the problem and explore various solution approaches. We will design and
evaluate fast failure detection and recovery techniques to minimize the disruptive effects of these
unexpected events. The direct impact of this research is the development of new algorithms and
techniques that will make scalable, high quality multimedia streaming over the Internet a reality. It
will develop key technologies for implementing new streaming software and impact every activity
involving access to multimedia contents over the Internet.
Keywords: Overlay network, multimedia streaming, multicast, failure detection, failure recovery
VI
Status | Finished |
---|---|
Effective start/end date | 5/1/05 → 4/30/08 |
Funding
- KY Science and Technology Co Inc: $68,062.00
Fingerprint
Explore the research topics touched on by this project. These labels are generated based on the underlying awards/grants. Together they form a unique fingerprint.