Neeraj Mittal wants to help computers collaborate more effectively.
Computer scientists used to focus on building enormous and powerful computers to solve enormous and complex problems. But they soon realized that even a monster computer had limitations, and the distributed system – multiple computers working together on the same problem – was born. Today, distributed systems are commonplace and doing important work on medical, mathematical and scientific problems.
But what if one computer in a system fails?
“My interest is in building distributed systems that can tolerate faults,” explains Dr. Mittal, an assistant professor of computer science. “If one or a few of them fail, then how do the others handle the failure? And what should they do so the system can tolerate the failure? This is most important for building systems containing millions of machines. It’s easier to design small-scale systems that work well, but the moment the system becomes large the performance degrades significantly.”
Dr. Mittal became interested in computers as a teenager in India. “I wanted to know how they worked and I wanted to build a computer myself,” he says.
He followed his interest into the Indian Institute of Technology in New Delhi, where he had his first hands-on experiences with computers. They were all he had hoped for and more.
“It was exciting,” he says. “They could do a lot of things I had not imagined. At that time the Internet was just starting, and I could spend hours on it.” As an undergraduate he also designed his first algorithms.
He started focusing on distributed computing in graduate school, and now he’s also doing research concerning the emerging field of cognitive radio as well as the security of wireless sensor networks.
He came to the United States at the age of 21 to pursue his doctorate at the University of Texas at Austin. “I chose UT Austin because of its strength in theory,” he says. “It also offered me a fellowship for two years.”
In Austin he zeroed in on his interest in distributed computing, in which he received his Ph.D. in 2002. Shortly afterward he joined UTD’s Department of Computer Science, where he’s now also pursuing his interest in what’s known as cognitive radio.
“This is something I started working on recently – the idea that radios operating over wide frequencies can sense the spectrum, find out what frequencies are available and start using them,” he says. “Initially, the main application will probably be military.”
He’s also researching ways to improve the security of wireless sensor networks.
“Communication over a wireless medium is very easy for an enemy to eavesdrop on,” he says. “You want to encrypt your communication to ensure that sensitive information is secure.”
Dr. Mittal is currently advising four graduate students, and he particularly enjoys teaching a class in advanced operating systems. “It’s in my core research area,” he says. “I know those topics very well. I know the current issues that are important right now and what people are working on.”
And, he notes, teaching improves his own research.
“It helps me understand the material better,” he says. “As you’re teaching, you’re learning.”
