Partha Kanuparthy

Ph.D. student
School of Computer Science

I'm a graduate student and a part of the Networking and Telecommunications Group. I'm advised by Constantine Dovrolis. I worked with Mostafa Ammar in 2006-2007, and with Ellen Zegura in fall 2006. My research interests include network inference and its applications. I was at Intel Labs Pittsburgh during Summer'10, working with Dina Papagiannaki, and Srini Seshan and Peter Steenkiste from CMU. I spent Summer'08 in the Systems and Networking group at Microsoft Research Cambridge (UK), hosted by Christos Gkantsidis, Thomas Karagiannis, and Peter Key. Prior to grad school, I was a member of the Research Division, Real-Time Collaboration Products, Oracle Corporation. I had spent four wonderful years at the Indian Institute of Technology Kharagpur, completing my undergrad in Mechanical Engineering with the batch of 2002.

I work on end-to-end inference of Internet path properties using active and passive methods. My research involves understanding these properties, statistical inference, and building tools and scalable, usable, systems for end-to-end inference. A clickable map of my work on network inference:

An end-to-end Internet path: my work on network inference.

DiffProbe: Detecting ISP Service Discrimination,
Partha Kanuparthy and Constantine Dovrolis
IEEE Conference on Computer Communications (INFOCOM) 2010, San Diego, CA, USA.
(Visit the DiffProbe page.)

DiffProbe won the Best Research Award at the FCC Open Internet Apps Challenge.

DiffProbe was presented at the CAIDA AIMS Workshop 2010.

Media: Some related articles:
"Google Launches M-Lab Measurement Tools," NetworkWorld, Jan. 2009.
"Google and Friends Release Net Neutrality Measuring Tools," Slashdot, Jan. 2009.
"Contest Shines Light on Broadband Providers' Tactics," PCWorld, Aug. 2011.
"FCC Contest Stimulates Development of Apps to Help Keep ISPs Honest," O'Reilly Media, Aug. 2011.

Please visit the ShaperProbe page. Hosted at Google MLab.

ShaperProbe: End-to-end Detection of ISP Traffic Shaping using Active Methods,
Partha Kanuparthy and Constantine Dovrolis
ACM SIGCOMM/USENIX/SIGMETRICS Conference on Internet Measurement (IMC) 2011, Berlin, Germany.

End-to-end Detection of ISP Traffic Shaping using Active and Passive Methods,
Partha Kanuparthy and Constantine Dovrolis
Extended Technical Report, May 2011.

ShaperProbe won the Best Research Award at the FCC Open Internet Apps Challenge.

ShaperProbe was presented at the CAIDA AIMS Workshop 2011 and at the Internet2 Joint Techs 2010.

Media: Appeared, amongst others, in:
"New 'ShaperProbe' tool detects ISP traffic shaping," Ars Technica, May 2011.
"Georgia Tech's ShaperProbe Detects ISP Traffic Manipulation," Slashdot, May 2011.
"Tool reveals ISP traffic-shaping," boing boing, May 2011.
"M-Lab Collaboration with ShaperProbe Shows That More Must Be Done to Help Consumers," New America Foundation, June 2011.
"PowerBoost: A Comcast Innovation for High-Speed Internet Users," Comcast Blog, August 2011.

Spectral Probing, Crosstalk and Frequency Multiplexing in Internet Paths,
Partha Kanuparthy, Constantine Dovrolis, and Mostafa Ammar
ACM SIGCOMM/USENIX/SIGMETRICS Conference on Internet Measurement (IMC) 2008, Vouliagmeni, Greece.

Spectral Probing was presented at the DIMACS/DyDAn Workshop on Internet Tomography 2008.

What can user-level active probing tell us about wireless links?

Can User-Level Probing Detect and Diagnose Common Home-WLAN Pathologies?,
Partha Kanuparthy, Constantine Dovrolis, Konstantina Papagiannaki, Srinivasan Seshan, and Peter Steenkiste
ACM SIGCOMM Computer Communications Review (CCR), Vol. 42, No. 1, January 2012.

An introduction to the Pythia project was presented at the CAIDA AIMS Workshop 2011 and the ESCC 2011.

Can we improve accuracy of DNS-based traffic engineering using online traffic measurements? Coming soon.

Hosted at Google MLab.
Over 60,000 runs as of May 2011.

Download access capacity (median Mbps) across the globe.

Indian Institute of Technology Kharagpur

partha [ ] cc . gatech . edu
Networking and Telecommunications Lab, College of Computing, Georgia Institute of Technology, Atlanta, GA 30332-0280.