NeTS: Small: Flexible Delivery of Streaming Video using Network-Aware MD-FEC Coding

Rensselaer Polytechnic Institute, Troy NY

Investigators

Abstract

Streaming video is envisioned to become the ?killer? networking application in the coming decade, and may determine whether the Internet and the current wireless technologies will live up to the explosive growth of multimedia demands of the next-generation network users. This project aims to develop coding and distribution mechanisms for transferring video over networks in a scalable and effective manner, which will be critical in meeting this challenge. In particular, the project investigates the design of a broad theoretical framework as well as specific mechanisms that would enable efficient delivery of streaming video through the use of video distribution networks (VDNs) and ?network-aware? Multiple Description Forward Error Correction (MD-FEC) coding. MD-FEC can be very effective in minimizing the overall distortion in video quality when delivering streaming video in a typical network environment that experiences time-varying bandwidth fluctuations and random packet losses. The project also explores how network coding can be integrated with MD-FEC coding to enhance live video streaming performance. This cross-disciplinary research is being done through a closely-integrated effort by an interdisciplinary team bringing together expertise in the areas of video coding and information theory; and network modeling, optimization and simulation. The results of this research project are expected to provide key insights on how streaming live video should be delivered for optimal performance over the Internet and last-hop wireless networks in the future. Proactive efforts are also being taken towards integrating the research findings into related graduate courses, and involving undergraduates in these research activities.