Information Trust Institute (ITI) Calendar

Back to Listing

TSS Seminar: Matt Caesar: "Towards a Trustworthy Internet Architecture"

Event Type
Information Trust Institute
2405 Siebel Center
Mar 11, 2009   4:00 pm  
Matthew Caesar, UIUC Department of Computer Science
Originating Calendar
Information Trust Institute (ITI) archival calendar


The Internet is an extremely large and complex distributed system, composed of tens of thousands of competing ISPs and hundreds of millions of potentially-misbehaving hosts. Concerns about the Internet's ability to meet ever-increasing demands on availability and performance in the presence of this complexity has led to a call to redesign the Internet's architecture. Towards this goal, we need a fresh approach to building network protocols with trustworthiness and robustness as primary design metrics. In this talk I'll overview three of our ongoing projects to address these problems. First, the complexity of Internet routers leads to implementation errors, or bugs, in protocol code. I'll describe our work on building bug-tolerant routing protocols, which prevent these bugs from affecting correctness of the network layer. Second, layer-2 networks are facing scaling challenges, from increasing deployments of data centers and large enterprises. I'll describe our work on designing a scalable Ethernet architecture, which allows construction of large layer-2 networks while maintaining Ethernet's self-configuring properties. Third, the convergence process of Internet routing protocols leads to outages and routing loops, and slows reaction to failures. I'll describe our work on designing a convergence-free routing protocol, which allows data packets to autonomously discover a working path without requiring up-to-date state in routers.


Matthew Caesar is an assistant professor in the Computer Science department at the University of Illinois at Urbana-Champaign. He completed his Ph.D. in Computer Science at the University of California, Berkeley. His research interests involve simplifying the management of distributed systems and networks through principles of self-organization and self-diagnosis, with an emphasis on networked systems, overlay networks, and the Internet. For more information, please visit

link for robots only