The university researchers who began construction on the internet some four decades ago never imagined the power their creation would have today. They toiled away in their labs quietly, and few outside cared.
That won’t be the case with a next-generation internet envisioned as an ultimate replacement for the current one.
Commercial and policy interests will likely play a bigger role this time, as researchers explore “clean slate” designs that scrap the internet’s underlying architecture to better address security, mobility, and other emerging needs.
Will the greater attention on these efforts ultimately be their undoing?
“The success of the internet can be largely credited to the fact that it began in a backwater,” said Jonathan Zittrain, a law professor affiliated with Oxford and Harvard universities. “It had the amazing advantage of not having to turn a profit. It didn’t need a business model.”
The bulk of the work is still being done in ivory towers, with grants from leading high-tech companies and government agencies.
Stanford University, for instance, has partnered with Cisco Systems Inc., Japan’s NTT DoCoMo Inc., Germany’s Deutsche Telekom AG, and other companies, though for now they are limited to advisory and sponsorship roles.
Bruce Davie, a Cisco fellow, said industry can take advantage of academia’s long-term vision, while giving feedback on what areas of research might actually be useful.
But commercial considerations are clearly in the minds of researchers.
Carnegie Mellon professor Hui Zhang said some of the work there surrounds building incentives for network operators to update systems and pass along data efficiently. Researchers are realizing they can’t simply rely on network operators’ altruism, a tenet in the original design.
Participants in a new network also could include law-enforcement officials, who are already demanding that internet service providers retrofit the existing network to ease wiretapping of internet-based phone calls. Governments around the world, including the United States, also could seek ways to block porn and politically sensitive web sites—and better identify those who distribute the forbidden.
“The more mature these ideas become, the closer they get to reality, I’m sure many stakeholders like that will come to the table,” said Larry Peterson, a Princeton professor who heads a planning group for an experimental network called GENI.
Building surveillance capabilities from the start could certainly cut costs, said Les Szwajkowski, a former FBI official who had worked on applying wiretap laws to new technologies. But he said engineers and other Americans shouldn’t worry.
“In theory this would be an excellent idea, but I think there are political issues to overcome,” Szwajkowski said. “There would be a reluctance to say you have an investigative agency at the table involved in a deep reworking of the internet.”
He said many in law enforcement share his view that any involvement should be limited to advising engineers rather than meddling in the details.
Guru Parulkar, incoming executive director of Stanford’s initiative following a tenure as the NSF’s clean-slate manager, said researchers recognize they must build any system with “the right balance of privacy and accountability,” leaving it flexible enough to adapt to wherever policy makers decide to draw the line.