Ram Durairajan wants to future-proof the Internet, and three new grants will help him do so.
The UO computer scientist has scored more than $2 million in funding from the National Science Foundation and $200,000 from the Internet Society Foundation. The money will support projects in his lab all aimed at making the internet more resilient against threats of all kinds.
“The internet is prone to a whole bunch of risks and attacks,” Durairajan said. For example, a major earthquake hitting Oregon could wipe out internet connections along the entire west coast and even further afield, with major economic implications.
And natural disasters such as hurricanes, floods and wildfires are only projected to increase in frequency and severity in the coming decades, due to climate change. Meanwhile, hackers are continually developing more sophisticated ways to maliciously attack internet infrastructure. They can bring down websites or intercept data, compromising privacy and security.
One of Durairajan’s new grants, from the National Science Foundation, will help address these threats by designing new ways to make networks more adaptable.
The internet has two main parts, Durairajan says: the topology and the traffic. The topology is how the network is built. It includes the pathways along which information can flow, and the way different nodes are linked together. In today’s internet, the topology is fixed. The traffic, the information that’s being sent, varies.
“Attackers can only control the traffic, not always the topology,” Durairajan said.
The NSF grant, in collaboration with researchers at Carnegie Mellon University and Boston University, will help him design a network with topology that can be programmed to change over time, putting up an extra barrier for bad actors. For would-be attackers, it would be like trying to rob an art museum with a constantly shifting floorplan.
Via a second grant, from the Internet Society Foundation, Durairajan will work on making the internet more resilient against natural disasters associated with climate change.
Just like an accident on a major highway during rush hour can back up traffic on surrounding roads for miles, a blow to key internet infrastructure could knock out connectivity across entire states or regions. Building more redundancy and risk awareness into the infrastructure could help lessen the impact from cascading risks, such as an earthquake followed by a tsunami or a flood.
And a third grant, also from the National Science Foundation, will help Durairajan make internet data science more collaborative. Working with colleagues at the University of California, Santa Barbara, the UO, and Niksun Inc., he’ll design ways to automatically classify types of internet traffic using machine learning.
That’s important for more efficient detection of normal versus malicious traffic and could help network operators catch performance issues and shut down attacks sooner. He’ll also develop and test systems to more effectively share that labeled and categorized data among users.
Durairajan hopes his work will call attention to the vulnerabilities in today’s internet and ultimately lead to policy changes that will improve it. But that requires cooperation from service providers and the government entities who regulate it.
“The internet is important for global common good, and improving it is a multistakeholder problem,” he said.
—By Laurel Hamers, University Communications