|
|
|
 |
Subhabrata Sen
AT&T Labs-Research
Friday, Mar. 4, 11:00am
LC 102, Brooklyn Campus, Polytechnic University
Abstract
The past few years have witnessed a dramatic increase in the number and
variety of applications running over the Internet and over enterprise IP
networks. An accurate mapping of traffic to applications is important for a
broad range of network management and measurement tasks including traffic
engineering, service differentiation, performance/failure monitoring, and
security. In the Internet, applications have traditionally been identified
using well-known default server network-port numbers in the TCP or UDP
headers. However this approach has become increasingly inaccurate because many
applications use non-default or ephemeral port numbers, or use well-known
port numbers associated with other applications.
In this talk, I shall present our exploration of 2 approaches to traffic
identification in the context of 2 real-world problems. First, we consider
the problem of accurately identifying P2P application traffic in real-time
on high speed links, and explore the use of content-based application
signatures. Our measurements show that our technique is accurate (less than
5% false positive and false negative ratios in most cases), and has good
scalability. Second, we consider the problem of Class-of-Service Mapping
for QoS in enterprise networks. We consider a solution framework for
measurement based classification of traffic for QoS based on statistical
application signatures. The method would be used off-line to form a set of
port, or IP address based rules for CoS assignment that would then be
applied on-line in the QoS implementation. Our evaluations indicate that the
technique has relatively low error rates, even for fine grain traffic
Bio:
Subhabrata Sen is a member of the Internet and Networking Systems Research
Center at AT&T Labs -- Research in Florham Park, New Jersey. He received a
Ph.D. in Computer Science from the University of Massachusetts, Amherst,
USA, in 2001. His research interests are in the field of Internet
technologies and applications, including network traffic measurement and
characterization, network security, network anomaly detection, peer-peer
systems and overlay networks, multimedia proxy services, and end-to-end
(operating system and network) support for streaming multimedia. Dr. Sen has
published over thirty four peer-reviewed journal, conference, and workshop
papers. He has served as a program committee member and reviewer for several
leading conferences and journals. He is currently serving on the organizing
committee of ACM SIGCOMM 2005 as Student Travel Grant Chair, and also
co-chairing the ACM SIGCOMM 2005 Workshop on Mining Network Data
(MineNet-05).
For more information please contact Nasir Memon (memon at poly.edu)