Professor Yi Cui has gotten recognition – and funding – for his novel idea that could give YouTube a run for its money.
The Vanderbilt assistant professor of computer science and computer engineering has won a prestigious National Science Foundation CAREER Award for his research into peer-to-peer networking to enable peer-to-peer multimedia streaming over the Internet.
He intends for his system to allow entrepreneurial Internet streaming video services to succeed without huge investments of capital on hundreds of gigantic computer servers. His plan will involve the computers of the subscribers, themselves, who will share video streaming data with other subscribers through an automated system.
That’s just the short version of the goal and a tiny glimpse of the challenge. Making it possible for centralized multimedia streaming servers to route video and audio signals through a vast network of subscribers, all with varying capabilities in terms of connection speed and processing power, is a feat that may well take the five years of research the NSF is sponsoring.
Current multimedia Internet streaming services, such as YouTube, must rely on centralized control by dedicated computer servers. The current system requires streaming multimedia data to individual computers through Internet connections, resulting in bottlenecks and slowed or interrupted delivery.
In Cui’s system, subscribers’ computers (“peers”) could become part of the multimedia streaming service network, allowing their available bandwidth to be borrowed to stream multimedia signals to other peer computers on the network.
This approach could reduce the price of the multimedia streaming service, help eliminate bottlenecks and distribute the electronic traffic more efficiently. It could also enable multimedia streaming services to enter the industry without investing as much capital in computer servers and bandwidth subscription.
“The NSF sponsorship will enable us to assess networked computers’ ability to transmit multimedia data, based on the customary use of the computer, the inferred bandwidth available to the computers, and a variety of customer usage patterns,” Cui said. He will test his system through the Open Source Teaching service, using his system to deliver multimedia educational materials freely.