ICTI News
João Barros, Faculty of Sciences at the University of Porto, to give talk on “Optimum Power Allocation for MIMO Systems with Arbitrary Inputs”
|
 |
Professor João Barros will be visiting Carnegie Mellon’s Pittsburgh campus the week of March 24, 2008. One of the highlights of Barros’ visit is a talk he will be presenting titled “Distributed Sensing: Fundamental Limits and Scalable Solutions.” The talk will be on March 25, 2008 at 4:30 in Porter Hall room B34 on the Carnegie Mellon campus. His talk is co-sponsored by the Information and Communication Technologies Institute (ICTI), part of the Carnegie Mellon | Portugal program, and Carnegie Mellon's Electrical and Computer Engineering department. |
Talk abstract:
Sensing, processing and transmitting data are arguably the key challenges of common nodes in a wireless sensor network, once they are embedded in a physical process unfolding in time and space. To process the data and transmit it reliably and efficiently over noisy links along the network, sensor nodes must run information processing algorithms capable of exploiting the natural correlation of the gathered data and of combating the impairments caused by noisy communication channels. In the first part of this talk, we will define reasonable models for the information sources and the communication channels and show how information theory offers powerful tools to help characterize the ultimate performance limits for this class of communication and computation systems. An important conclusion to be drawn from our results is that for a large (and most relevant) class of sensor networks, separate data compression and error correction codes provide an optimal system architecture. The proofs also offer hints on how to construct practical algorithms for distributed compression and joint inference of correlated data collected by hundreds of sensor nodes. In the second part, we show that, even if we use the simplest possible coding scheme at the sensor nodes (a scalar quantizer and a modulator), implementing the optimal decoder for minimum mean square estimation (MMSE) is unfeasible – its complexity grows exponentially with the number of nodes. Using factor graphs and belief propagation tools, we are able to present scalable alternatives, which offer a precise trade-off between decoding complexity and end-to-end distortion. Finally, we discuss how these ideas lead to natural solutions for low-complexity distributed quantization, source-optimized hierarchical clustering, and estimation of functionals over noisy channels. Joint work with Gerhard Maierbacher and Michael Tuechler.
Biography:
João Barros received his undergraduate education in Electrical and Computer Engineering from the Universidade do Porto (UP), Portugal and Universitaet Karlsruhe, Germany, until 1999, and the Ph.D. degree in Electrical Engineering and Information Technology from the Technische Universitaet Muenchen (TUM), Germany, in 2004. After his doctoral research on network information theory and joint source and channel coding, João Barros joined the faculty of the Department of Computer Science at the Universidade do Porto, where he founded the Networking and Information Processing Group of the Instituto de Telecomunicações. The focus of his research lies in the general areas of information theory, communication networks and data security. In 2003, Dr. Barros received a Best Teaching Award from the Bavarian State Ministry of Sciences, Research and the Arts. In 2002 and 2003, he spent six months as a Fulbright scholar at Cornell University, where he worked on fundamental limits of wireless sensor networks. João Barros has served on several Technical Program Committees, including WiOpt 2008, ISIT 2007, IS 2007, Globecom 2007 and SSI 2006. Since July 2006, he serves as Secretary of the Board of Governors of the IEEE Information Theory Society. From January through August 2008, João Barros is on sabbatical leave at the Laboratory of Information and Decision Systems of the Massachusetts Institute of Technology.