Computer Science Colloquia
Monday, January 23, 2012
Nathanael Paul, Oak Ridge National Labs / University of Tennessee
Host: Dave Evans
Rice Hall, Room 242, 3:30 pm
Computer Science Department Colloquium
Trustworthy Systems from (Un)trusted Components
Many trustworthy systems today must be engineered to run with untrustworthy components. This talk will focus on challenges in building trustworthy systems for voting, embedded systems (i.e., health care), and peer-to-peer networks. In voting, jurisdictions have experienced challenges with punch-card voting systems and electronic voting systems, and many have returned to using paper ballots. However, rather than foregoing the benefits of an electronic voting system, we have designed an electronic voting system that emphasizes a systemic view and defense-in-depth. In health care, recent work has shown that portable medical device systems are susceptible to security breaches that can impact patient health. By augmenting the system with more data from the patient’s environment, we are able to create a forensic system that provides increased auditability and security. In peer-to-peer networks, recent research has shown that it is possible to track an IP address to less than a mile from its geo-physical address. We will discuss the lack of privacy in a peer-to-peer system and summarize our approach of pinpointing the exact physical location to find users taking part in suspected illegal activity.
Nathanael Paul is a research scientist leading the Center for Trustworthy Embedded Systems at the Oak Ridge National Laboratory, and holds a joint appointment as an associate professor with the University of Tennessee. His research focuses on security and privacy for embedded systems. Current projects include creating trustworthy insulin pump systems (Univ. Tenn.) and evaluating the privacy and anonymity in large-scale P2P networks (Oak Ridge National Laboratory). As part of the medical device security work, Paul serves as a co-director for the Medical Device Security Center (secure-medicine.org). His work has been widely cited in major media news outlets including MSNBC and CNN. Paul received a PhD in Computer Science from the University of Virginia in 2008.