UA-linked effort aims to retool workings of Net

Joint project to redesign the retrieval of online info, speed it up in process
2010-12-06T00:00:00Z UA-linked effort aims to retool workings of NetVictoria Blute Arizona Daily Star Arizona Daily Star
December 06, 2010 12:00 am  • 

The Internet lets us watch television shows, connect with friends and even have groceries delivered if we don't want to brave the store. It's used for just about everything - but that doesn't mean it's perfect.

That's where Beichuan Zhang comes in. The University of Arizona computer scientist is part of a group of researchers who want to change the Internet's architecture. He's working with colleagues from nine other institutions on the Named Data Networking project to restructure the building blocks of the Internet, one of four projects the National Science Foundation is funding for up to $8 million.

The project's goal is to move away from the client-server relationship on which the Internet now relies. The way the Internet communicates information now means the data's location is more important than the data alone.

It's why, for example, trying to access a popular video online can be frustrating when many other people also are trying to watch it. They're each opening individual connections with the video's server - requests that a single server has to respond to that can result in bottlenecks.

The architectural change has a lot to do with performance, security and usability, Zhang said.

Instead of a router that transmits a video clip a number of times, the group wants to create a system that will place a unique marker on the packet of information where the clip is. This will allow the first user in a given location to download that video, at which point a router can forward that same clip to everyone else in that location who requests it - which means the clip will have to travel only once, cutting down on the waiting time.

The way the Internet is used today is completely different from its original intended use - scientific and military networking, Zhang said. And that's the big problem.

The Internet's architecture was perfect for the original use, but as the Web grew more popular and expanded to a larger audience, problems such as scalability and security came up, Zhang said.

"The Internet was built in the early '70s, and the fundamental architecture hasn't changed much," Zhang said. "However, almost everything else has changed. The applications have changed. We use the Internet for online banking, Facebook and YouTube. It's totally different."

Scientists have tried to make small improvements in the past, but this is the first time a large group has come together with this consensus, Zhang said.

Project manager James Thornton, who works at the Palo Alto Research Center, said this is the first model that considers memory and storage as part of how the Internet works.

But if the Internet is really for public use, when will the public have access to the redesign ideas? Zhang isn't sure yet.

"It's a question I really want to be able to answer, but I have no idea," he said.

However, all the research is public, and the results will be available in an accessible Web prototype so that interested parties can try it, he said.

Thornton said the Palo Alto Research Center already has produced initial open-source software, but the final product will be expanded from that base.

At this point, the software is aimed at the research community, and it isn't offering much in the way of public use, he said.

But he also said that as the process advances, experimental applications will be built and may be released "for people to play with." He said companies might begin to do things within their own networks with potential for large-scale use within the network as a whole. In that case, the general public may experience the benefits of the research without realizing it, he said.

Zhang said the scientists intend to develop a prototype that displays the benefits and feasibility of their ideas by the time the grant funding ends in three years. He said that if the project is successful, the group hopes to attract more collaborators.

"When I post a query to say: 'I want this information, Internet. Can you get this?' - it doesn't matter where it comes from," he said. "The content can come from San Jose, New York, Phoenix, Tucson or anywhere. I don't care, as long as I get it and can verify it's exactly what I want. This will change things in a totally different way. If we can realize this idea, it will make our information retrieval much more efficient."

Contact NASA Space Grant intern Victoria Blute at vblute@azstarnet.com

Copyright 2014 Arizona Daily Star. All rights reserved. This material may not be published, broadcast, rewritten or redistributed.

Follow the Arizona Daily Star

Featured businesses

View more...
Get weekly ads via e-mail