Research interests

My current research interests lie mainly in the areas of information theory and statistical signal processing related to communication theory. In the context of wireless network communication, I am currently investigating statistical inference algorithms to achieve joint compression and network coding of correlated sources.

Research experiences

Jan 2011 - Feb 2012 :     post-doc fellow, EECS Department, University of Michigan, Ann Arbor, MI

advisor: Prof. A. Hero
research project: Correlation detection in high dimensional data sets.
abstract: When the number of variates in a data set largely exceeds the number of available samples, classical inference techniques cannot be applied to the discovery of the correlation structure of the random vector. I worked on an hypothesis test for correlation detection for elliptically contoured variables.

Sep 2006 - Dec 2010:     PhD student, LSS

advisor: Prof. M. Kieffer
research project: Theory and practice of Wyner-Ziv coding with uncertain side information quality.
abstract: When exploited in the design of practical applications, Wyner-Ziv coding proves not to be robust with respect to the uncertainty in the distribution of the correlation noise, which can dramatically decrease performance. I considered alternative statistical models for the correlation channel (Gaussian mixtures), providng rate-distortion analysis and design strategies for practical coding schemes in absence of feed-back channel.


Jan 2009 - May 2009:    visiting PhD student, DEIS, Università di Bologna, Italy

advisor: Prof. M. Chiani
research project: Theory and practice of Wyner-Ziv coding with uncertain side information quality.

Jan 2007 - June 2007:  visiting PhD student, DEIS, Università di Bologna, Italy

advisor: Prof. M. Chiani
research project: Theory and practice of Wyner-Ziv coding with uncertain side information quality.


Résumé