Enabling Future Multimedia Information Systems to Leverage Next-Generation Networking Technologies

本研究的目的是开发一套技术， 低成本的视频传输。 我们探索一种新的多播版本 所谓的动态组播，一个聪明的方式来提供视频到家庭， 2-阶段服务模型，以及一种有效的技术，用于传播视频点播卫星称为摩天大楼广播。 - 动态组播是为了解决网络I/O瓶颈而设计的， 阻碍服务器充分利用网络容量。 通过这种技术，系统能够支持更多的同时 多播比服务器的物理能力。 - 摩天大楼的设计是为了利用 卫星. 在这种环境中，由于带宽要求是 与视频数量成正比，而不是用户数量，我们有一个 非常可扩展的技术。 在成本方面，这种方法 基本上不需要布线，这使得它非常吸引人。 - 第三种技术，即2-Phase Service Model，适用于 在视频回放需要比物理比特率更高的比特率的情况下 中能维持。 家庭用户可以从这项技术中受益。 它 为搜索高分辨率视频提供了非常有效的环境 图书馆从家里 为了证明这些想法的可行性，我们实现了三个 视频点播系统 我们还进行实验，以评估这些技术使用真实的视频数据的性能。 有关此计划的详情，请浏览网页 http://www.dsg.cs.ucf.edu/NSF/nsf.html.

The objective of this research is to develop a suite of techniques to allow video distribution at low cost. We explore a novel version of multicast called Dynamic Multicast, a clever way to deliver videos to homes called 2-Phase Service Model, and an efficient technique for disseminating videos on demand over satellites called Skyscraper Broadcasting. - Dynamic Multicast is designed to address the network-I/O bottleneck which hinders the server from taking full advantage of the network capacity. With this technique, the system is able to support many more simultaneous multicasts than the server is physically capable of. - Skyscraper is designed to exploit the unique broadcast capability of satellites. In this environment, since the bandwidth requirement is proportional to the number of videos, not the number of users, we have a very scalable technology. In terms of costs, the fact that this approach requires essentially no wiring makes it very appealing. - The third technique, 2-Phase Service Model, is intended for environments where the video playback requires a higher bit rate than the physical medium can sustain. Home users can benefit from this technology. It offers a very effective environment for searching high-resolution video libraries from home. To demonstrate the feasibility of these ideas, we implement three video-on-demand systems. We also conduct experiments to assess the performance of these techniques using real video data. For more information on this project, please visit the web page at http://www.dsg.cs.ucf.edu/NSF/nsf.html.