Purdue forges top cyber research center
Purdue University is home to the nation’s first computer science department, and since 1998, it has hosted the Center for Education and Research in Information Assurance and Security (CERIAS), a multidisciplinary center that supports work in software and hardware security, security policy and ethics, risk management, computer crime, and information warfare.
A land-grant college founded in 1869 to provide education in agriculture and engineering and whose team name is the Boilermakers, Purdue might seem to be an unusual setting for pioneering programs in computer science and security research. But high tech has deep roots at Purdue.
Security is stuck in a rut
“Purdue has been a major science and engineering school from its formation,” said Eugene Spafford, professor of computer science and director of CERIAS (pronounced “serious”).
During its first century, Purdue’s schools of chemical, mechanical and electrical engineering provided a large percentage of the nation’s engineers, and in the 20th century, it became a national leader in aeronautical engineering. Neil Armstrong and Eugene Cernan, the first and last men to walk on the moon, are among 22 Purdue grads who became astronauts.
The emergence of computer science as an academic discipline at the university in the 1960s was a natural outgrowth of its background in advancing scientific technology. “Being at the cutting edge, the school got grants to do computation on campus,” Spafford said. And as computers became available, a need developed for trained grad students who could operate them. “There was a demand for the classes,” which led to degree programs.
The country’s first computer science department was established at Purdue in 1962. It initially offered graduate degrees with some undergraduate service courses in programming, and an undergraduate bachelor's degree in science was added in 1967.
As computers became ubiquitous and networked, the need for security began to be apparent, and CERIAS has emerged at the leading edge of computer security research.
Since its creation 13 years ago, the center has assembled a faculty of nearly 80, drawn from 20 departments on the West Lafayette, Ind., campus, who collectively have authored nearly two-dozen books on computer and information security. CERIAS was also one of the National Security Agency’s original National Centers of Academic Excellence in information assurance education.
Spafford, its founding director, has served on the President’s Information Technology Advisory Committee and has been an adviser to the National Science Foundation. Spafford’s roots in cybersecurity run deep. He reverse-engineered the Morris worm in 1988, one of the first Internet worms. He helped develop the Computerized Oracle and Password System, an early open-source security scanner, with student Dan Farmer, and the Tripwire compliance automation tool with Gene Kim, who went on to found Tripwire Inc. in 1997.
Spafford came to Purdue in 1987 from the Georgia Institute of Technology, where he completed his graduate work and developed his interest in computer security. While at Georgia Tech, he did research on fault-tolerant network machines for distributed computing — a forerunner of cloud computing — and, as part of that work, became involved in software testing and debugging. He also was an administrator for 11 years on Usenet, a distributed Internet discussion system of threaded newsgroups that predated the World Wide Web.
While monitoring one of the Usenet servers in November 1988, he discovered it had been infected by the Morris worm, and his analysis of it became a focal point for efforts to halt this early malware.
Legacy security demand
In the 1980s, there was little pursuit of computer security as an academic study because it had been mathematically demonstrated that complete security was impossible. But by the 1990s, the task of securing legacy systems had made computer security a practical commercial field with funding available, and Spafford had achieved a degree of prominence because of his work with the Morris worm.
“It was an area really ripe for me to do things,” Spafford said. He and Samuel Wagstaff established Computer Operations, Audit, and Security Technology, a software security lab in 1991. The lab was part of Purdue’s Computer Science Department, with initial funding from Sun Microsystems, Schlumberger, Bell-Northern Research — which later became Nortel — and Hughes Laboratories.
COAST grew during the next six years to become the country’s largest computer security research group, with an annual research budget of $1 million. But Spafford chafed at the lab’s limited scope of software and network security for legacy systems. “I was interested in things like the ethics and security of privacy,” he said. He wrote an early article on the ethics of hacking. “That didn’t sit well with my colleagues here.”
But when he received an invitation to return to Georgia Tech to set up a security research center in 1998, Purdue gave him permission to set up his own center there. Within a few years, CERIAS absorbed COAST and also drew on the work of faculty in other departments to investigate all facets of security. Seed funding came from a $4 million grant from the Lilly Endowment.
Despite the academic and financial support, CERIAS has remained a small, low-overhead operation with a small full-time staff. Spafford is a full-time faculty member in the computer sciences department in addition to his work with the center. “For running the center, I get the equivalent of one week’s salary,” he said.
Most of the Lilly funding went to support small research programs in other departments throughout the university.“We spent it all in about three years,” Spafford said. “But in that time, we funded about 80 projects on campus, hired four faculty and supported about a dozen students in their graduate studies.”
This expanded interest in the work spurred additional funding, which has allowed CERIAS to grow. It eventually replaced COAST, and in 2002, it established an interdisciplinary master's degree program in departments across the university. That was expanded in 2008 to a doctorate program, which has so far awarded one degree with three more in progress. The center has brought an estimated $40 million to $50 million in research funding to Purdue during the past decade and through programs in all of the associated groups has produced about 100 doctorates in computer science.
According to NSF, Purdue ranked No. 14 in the nation in physical science doctorates awarded in 2009, right behind Harvard, with 118, including four doctorates in computer and information science. Nationwide, there were 1,364 computer science doctorates awarded in the United States that year.
Despite its school ties, “the center is largely self-supporting,” Spafford said. “We exist on outside funding.” With the original Lilly grant long spent and corporate donations of equipment and money shrinking because of the economy, fundraising is becoming a large part of the director’s job. “That can be a challenge,” he said.
And as a practical issue, it also affects the work being done at CERIAS. “If we had a big bundle of discretionary money, we would be able to set the agenda” for security research. In the real world, it is the government agencies and corporate sponsors funding the work that call the shots. But CERIAS still works to maintain its focus on forward-looking technology and policy rather than working exclusively on securing legacy systems.