Grassroots Privacy Techniques in Crowdsensing Applications

29 Nov
Monday, 11/29/2010 11:00am to 12:00pm

Raghu Ganti
Messaging and Event Systems

Computer Science Building, Room 151

Faculty Host: Deepak Ganesan

Embedded sensing devices such as smartphones, music players, and GPS devices are now pervasive and part of our everyday lives. The availability of these devices will soon result in the rise of people centric sensing, where devices owned by individuals are the primary source of sensor data. An important category of applications are those that compute, model, or map common phenomena of mutual interest from sensor data generated by a group of individuals. For example, individuals share their location and speed data from their daily commutes to map traffic patterns. A flipside to sharing sensitive sensor data, such as location information is that the privacy of an individual is compromised and can result in serious breaches, such as shown by the famous website. In this talk, I will present our approach to privacy that preserves the privacy of an individual sharing sensor data and at the same time enables one to compute, model, or map common phenomena of mutual interest accurately. Our approach relies on perturbing or transforming the sensor data at the individual's end through the addition of appropriate noise or by transforming the sensor data itself into a different form. I will illustrate with a few applications that we have developed which utilize the above approach to privacy.


Raghu Ganti is currently a Research Staff Member at IBM's T. J. Watson research center. He is currently working on developing an infrastructure for enabling mobile crowdsensing applications. He received his M.S. (2006) and Ph.D. (2010) degrees from the Department of Computer Science at the University of Illinois, Urbana-Champaign under the guidance of Prof. Tarek Abdelzaher. During his Ph.D, he worked on developing a generic set of tools for human centric sensing, which included privacy preservation, activity identification, and modeling community phenomena. He was a graduate student at the Department of Computer Science, University of Virginia from 2003 to 2005. He received his B.Tech. from Indian Institute of Technology, Madras in 2003 in Computer Science and Engineering. His major research interests are in the areas of mobile crowdsensing, cyber-physical systems, and data mining. He is the recipient of the Siebel Scholar fellowship, Class of 2010, which is awarded annually for academic excellence and demonstrated leadership to the top 80 graduate students across the world's leading graduate schools.