04-17

Bots, Bubbles, and Bottlenecks: Safeguarding the User’s Internet Experience

[[{"fid":"443","view_mode":"embedded_left","fields":{"format":"embedded_left","field_file_image_alt_text[und][0][value]":"Nick Feamster","field_file_image_title_text[und][0][value]":"","field_file_caption_credit[und][0][value]":"%3Cp%3ENick%20Feamster%3C%2Fp%3E%0A","field_file_caption_credit[und][0][format]":"full_html"},"type":"media","attributes":{"alt":"Nick Feamster","height":375,"width":250,"class":"media-element file-embedded-left"},"link_text":null}]]A user's experience on the Internet rests in the hands of a large and increasingly diverse set of stakeholders. Internet service providers and content providers point fingers at one other about performance problems.  Miscreants launch attacks against both other users and the Internet infrastructure itself.  Content providers continually engage in practices to "personalize" what we see and when we see it.   Many governments aim to control its citizens' access to information, while activists design circumvention tools.   Safeguarding the user's Internet experience requires both gathering empirical network measurements to detect threats (typically in the absence of any "ground truth") and developing large-scale systems to mitigate them.  In this talk, I will present three classes of safeguards against different threats to the user's Internet experience: (1) technologies to characterize and improve performance in the Internet's "last mile", including a worldwide deployment of home routers in around 200 home networks and ongoing studies with the Federal Communications Commission; (2) methods for lightweight and fast detection of message abuse, such as spam, that have since been widely adopted by industry; and (3) defenses against against information manipulation attacks, a new class of attacks against personalization algorithms.  I will also discuss other such threats and how networking can draw from other disciplines to tackle them.

Nick Feamster is an associate professor in the College of Computing at Georgia Tech. He received his Ph.D. in Computer science from MIT in 2005, and his S.B. and M.Eng. degrees in Electrical Engineering and Computer Science from MIT in 2000 and 2001, respectively. His research focuses on many aspects of computer networking and networked systems, with a focus on network operations, network security, and censorship-resistant communication systems. In December 2008, he received the Presidential Early Career Award for Scientists and Engineers (PECASE) for his contributions to cybersecurity, notably spam filtering. His honors include the Technology Review 35 "Top Young Innovators Under 35" award, the ACM SIGCOMM Rising Star Award, a Sloan Research Fellowship, the NSF CAREER award, the IBM Faculty Fellowship, the IRTF Applied Networking Research Prize, and award papers at the SIGCOMM Internet Measurement Conference (measuring Web performance bottlenecks), SIGCOMM (network-level behavior of spammers), the NSDI conference (fault detection in router configuration), Usenix Security (circumventing web censorship using Infranet), and Usenix Security (web cookie analysis).