On the Multiterminal Interactive Lossy Source Coding Problem

Add to your list(s) Download to your calendar using vCal

• Dr Pablo Piantanida, SUPELEC, France
• Wednesday 17 September 2014, 11:00-12:00
• EEE Department, Room 503.

If you have a question about this talk, please contact Patrick Kelly.

Abstract: Efficient distributed data compression may be the only way to guarantee acceptable levels of performance, e.g., anomaly detection, when energy and link bandwidth are severely limited as in many real world sensor networks. The distributed data collected by different nodes in a network can be highly correlated and this correlation can be exploited at the application layer, e.g., for target localization and tracking or anomaly detection. In such cases cooperative joint data-compression can achieve a better overall rate-distortion tradeoff than can do independent compression at each node. Interaction among nodes may take place via distributed/successive refinement source coding, where nodes exchange (interactively) data among themselves over a given number of communication rounds.

In this talk, we first introduce the multiterminal interactive lossy source coding problem. This scenario consists of a network composed of multiple modes which can interact through rate-limited (error free) links. Each node measures the realization of a memoryless source and is required to reconstruct the sources from the other terminals with a fidelity criterion. Nodes are allowed to interact by interchanging information over a finite number of communication rounds. When the information exchange phase is over, the nodes try to reconstruct the realization of the sources at the other nodes using all decoded descriptions. We first derive a (rather involved) inner bound to the general rate-distortion region of this problem. It is shown that this inner bound contains most of the existent achievable rate regions in the field. Then, we prove that our inner bound provides the optimal rate-distortion regions for several relevant source coding settings, including cooperative distributed source coding problems.

Joint work with Alfred O. Hero III (University of Michigan) and Leonardo Rey Vega (University of Buenos Aires).

Short biography: Pablo Piantanida received the B.Sc. and M.Sc degrees (with honors) in Electrical Engineering and Mathematics from the University of Buenos Aires (Argentina), in 2003, and the Ph.D. from the Paris-Sud University (France) in 2007. Since October 2007 he has joined the Department of Telecommunications, SUPELEC , as an Assistant Professor in network information theory. His research interests include multi-terminal information theory, Shannon theory, cooperative communications, physical-layer security and distributed source coding.

This talk is part of the Featured talks series.

This talk is included in these lists:

• Andrea Picciau's list
• COMMSP & CP list
• COMMSP Seminar
• EEE Department, Room 503
• Featured talks
• SAP Talks

Note that ex-directory lists are not shown.