Department of Electrical and Computer Engineering (archive)

Distributed Inference in Sensor Networks: Topology and Tradeoffs Jan. 22

 

Professor José M. F. Moura, Department of Electrical and Computer Engineering, Carnegie Mellon University, will present a seminar entitled Distributed Inference in Sensor Networks: Topology and Tradeoffs on January 22, 2006 from 1:00-2:00 PM in Alliant Energy - Lee Liu Auditorium, Howe Hall.

Abstract:
Sensor networks, which can instrument large areas with many inexpensive multimodal sensors, have the potential to collect vast amounts of data. Extracting relevant information from this data locally to avoid massive data transfers and localizing inter-sensor communication are important issues to reduce power consumption, while still achieving good performance. Distributed network architectures avoid centralized decision and do not overburden a few sensors with communication and processing, rather distributing across all sensors traffic flow and processing. By network topology, we mean the graph that supports the communication interchange among sensors. We consider the problem of optimizing the topology of the sensor network for distributed inference, subject to a constraint on the maximum number of available communication channels among sensors. We convert this into a spectral graph problem, i.e., what is the graph that maximizes the ratio between two graph spectral parameters – the graph algebraic connectivity and the maximum eigenvalue of the graph Laplacian. We present results for different versions of this problem, including when the communication costs vary among sensors and when the sensors may fail at random times (communication errors).

Biography of Professor Moura
José M. F. Moura is a Professor of ECE at CMU and founding co-director of CenSCIR, the Center for Sensed Critical Infrastructure Research. In 2006-07, he is on sabbatical at MIT as a Visiting Professor of EECS. He holds MSc, EE, and D.Sc. EECS degrees from MIT and an EE degree from Instituto Superior Técnico (Lisbon, Portugal). His interests are in algebraic and statistical signal/ image processing. His current research includes distributed decision in sensor networks, time reversal imaging, SPIRAL, an intelligent compiler for signal processing transforms, and bio-imaging.

He  serves as President Elect for the IEEE Signal Processing Society (SPS), on the Editorial Board of the IEE Proceedings and of the ACM Journal, on several IEEE boards, and served  or serves on the steering committee of two major Conferences: the IEEE International Symposium on BioImaging (ISBI) and the ACM/IEEE International Symposium on Information Processing and Sensor Networks (IPSN). He was the Editor in Chief for the IEEE Transactions on Signal Processing. He is a Fellow of the IEEE, a Fellow of AAAS, and a corresponding member of the Academia das Ciéncias of Portugal. He received the IEEE 3rd Millennium Medal and the IEEE SPS Meritorious Service Award.

To view a video of Moura's lecture, click here.

---