University of Hawaii

Department of Electrical Engineering

Democratizing Content Distribution

Date: 2008-07-11
Time: 10:30
Location: HOLMES 389
Speaker: Prof. Michael J. Freedman


In order to reach their large audiences, today's Internet publishers primarily use content distribution networks (CDNs) to deliver content. Yet the architectures of the prevalent commercial systems are tightly bound to centralized control, static deployments, and trusted infrastructure, thus inherently limiting their scope to ensure cost recovery. 

This talk first outlines a number of techniques (and the resulting  systems) towards highly-scalable cooperative content distribution. We provide three central mechanisms of CDNs---content discovery, server selection, and secure content transmission---by federating large  numbers of unreliable or untrusted hosts. These ideas have been implemented, deployed, and tested in production systems (CoralCDN, OASIS) currently serving several terabytes of data to more than a million people every day.

Yet to realize cooperative content distribution writ large, an ideal CDN system should also promote peer participation, recognize network operator interests, and support server provisioning.  Time permitting, we discuss some ongoing work that, taking a market-based pricing approach to resource management, helps satisfy the goals of all
participants in the system, while still efficiently allocating resources across multiple files, avoiding resource congestion, and leveraging locality.
Michael J. Freedman is an assistant professor of computer science at Princeton University.  Prior to that, he received his Ph.D. in computer science from NYU and his S.B. and M.Eng. degrees from MIT. His research interests broadly focus on distributed systems, security, networking, and cryptography.  He developed and operates CoralCDN, a peer-to-peer content distribution network, and OASIS, an open anycast service, which serve more than a million users daily.  Other research
has included system fault tolerance and fault detection, privacy-preserving operations on datasets, secure enterprise network architectures, IP geolocation and intelligence, secure distributed file systems, and various anti-censorship, privacy-enhancing, and
anti-spam systems.