Intel Research Seminar

DATE: Thursday, October 6, 2005
TIME: 12.00 noon – 1.00 pm
PLACE: Intel Seminar (CIC Suite 410)

Zhen Xiao
AT&T Labs – Research

Optimizing Internet Content Delivery Systems -- A View from the Inside

With the enormous growth of Web and multimedia traffic on the Internet, various techniques have been proposed to optimize resource utilization and to improve quality of service for Internet applications.  The techniques include Web caching, streaming proxies, content delivery networks, and utility computing. While many of these techniques can be very effective, it remains unclear how and when they are best used.

In this talk, I will first present two comprehensive performance studies of the commercial workload collected from a large number of Web sites hosted by a major ISP and that collected from a large group of home users connected to the Internet via a well-known cable company.  Our studies have found significant inefficiency in the existing practice of Internet content delivery.  For example, most Web sites do not utilize the cache-control features of the HTTP 1.1 protocol and make indiscriminate usage of cookies.  In addition, we also found that the majority of multimedia content is still delivered via downloading from Web servers instead of via streaming.

Then I will present our tools/systems for improving the current practice in content delivery.  I will describe the Cassandra toolkit that can help content providers improve the performance of their Web sites.  Cassandra combines performance and behavioral data with extensible simulation architecture to identify content delivery problems and predict optimization benefits.  I will show how a hosting service provider can use Cassandra to estimate the benefit of using a content delivery network.  I will also describe our utility computing work that pushes data and computing from centralized sites to resources available at the edge of the network.  In our system, a content provider can deploy an application on a single computer anywhere in the network, and the system will automatically redeploy the application on a set of distributed servers as required by user demand and distribute user requests among those application replicas based on load and proximity.  Finally, I will give an overview of some other research I did at AT&T and Cornell.

Zhen Xiao received his Ph.D. from Cornell University in January 2001.  He is currently a senior technical staff member of the Distributed Systems Research Department at AT&T Labs -- Research in Florham Park, NJ.  His research interests include Web technologies and content delivery, multimedia and streaming, peer-to-peer systems, security and dependability, and reliable multicast.  He has published about 20 papers in leading journals and conferences, including ACM Transactions on Computer Systems, USENIX, WWW, IMC, Infocom, DSN, and ICDCS.  He has also served on four NSF panels and various program committees, including WWW, Infocom, and ICDCS.  He is the recipient of the AT&T Labs “Research Excellence Award” (2002) for his outstanding achievement on optimizing Internet content delivery.

For more information, please visit:

For Further Seminar Info:
Contact Kim Kaan, 412-605-1203, or visit

SDI Home: