UH is one of 10 schools selected to participate in the National Science Foundation’s Global Environment for Network Innovations (GENI) project.
“NSF started this maybe 10 years ago as an initial brainstorming phase, and then five years ago, it started funding it,” Engineering associate professor Deniz Gurkan said.
“It’s a clean-slate Internet design project.”
There are a limited number of protocols used to actually relay data information, with the primary one being Internet protocol (IP). GENI’s end goal is to redesign the Internet for the modern user by developing a better protocol to replace or improve upon this.
“If we design protocols now with the knowledge that we have now about Internet and how it is used, how would we design them?” Gurkan said.
“Forget about IP, forget about TCP, how would you have done it now that you know where it ended up? It’s going to revolutionize (how we use the Internet).”
Gurkan has been working on the project since 2008 and helped to develop a test bed where they can run potential ideas for improved protocols. The test bed has to run IP, but it is capable of running other protocols as well.
“The initial phases have been only building the infrastructure,” Gurkan said.
“I actually helped with that with the resources, collaborations around the Houston area. Many universities around the nation are helping also.”
The main goal is to build an underlying research-based infrastructure for the protocol where users can choose to opt-in at their own risk
It could potentially affect not just how users access the Internet from home devices, but also their mobile connections.
“They can just use that as a data network instead of subscribing to their cellular data network. So whatever form of networking they do, they do it on this network,” Gurkan said.
“But they sign an agreement with us saying that we can use their data to design the protocols based on real usage. That’s the experimentation part of the GENI, which is the most exciting part.”
As for the issues with IP that led to this reconsideration of how the Internet functions, Gurkan said that it’s hard to pinpoint just one specific problem.
“It’s a multi-dimensional set of issues. I can’t say IP is the reason we need to do a clean-slate design of the Internet,” Gurkan said.
“IP is designed as a layered architecture. All of these layers were designed to be isolated from each other. With the bandwidth being so large, processing powers and the cost being low enough, we may not need those layers anymore.”
Rather than assign an IP address to each user and have that be the only point of reference for locating them, Gurkan poses the possibility of developing a protocol that classifies users by something else.
“Not an address, but an interest level. That’s how all of the content-providers are doing their work now,” Gurkan said. “They have a server forum that’s located all over the world and they find a server that can serve you in the fastest way.”
This project plans to expand to 30 to 40 additional campuses within the next two years, according to Gurkan. A White House initiative has also begun called US Ignite, which encourages users to propose applications to NSF to receive funding.