IEEE Winnipeg Section


UofM ECE Department Seminar


Clustering, Estimation, and Mobility in Wireless Sensor Networks: Accounting for Energy Costs and Communication Constraints


Monday, November 21, 2011 from 3:30pm


E2-361 EITC Building, University of Manitoba Fort Garry Campus


Edward J Coyle
Georgia Institute of Technology


In networks of wireless sensors that have been organized into a hierarchy of clusters, each sensor collects noisy observations of the environment, quantizes these observations into a local estimate of finite length, and forwards them through one or more noisy wireless channels to the Cluster Head (CH). We focus in this talk on the processing a CH must perform to optimally fuse the local estimates it receives from the nodes in its cluster. The measurement noise seen at each sensor is assumed to be zero-mean and have finite variance and each wireless hop is modeled as a Binary Symmetric Channel (BSC) with a known crossover probability. A novel scheme is proposed that uses dithered quantization and channel compensation to ensure that each sensor’s local estimate received by the CH is unbiased. The CH fuses these unbiased local estimates into a global one using a Best Linear Unbiased Estimator (BLUE). Analytical and simulation results show that the proposed scheme can achieve much smaller mean square error (MSE) than two other common schemes while using the same amount of energy. The sensitivity of the proposed scheme to errors in estimates of the crossover probability of the BSC channel is studied by both analysis and simulation. Algorithms are provided that can determine: (1) The minimum energy required for the network to produce an estimate with a prescribed error variance; (2) The optimal allocation of this energy amongst the sensors in the multi-hop network; and (3) The effect of mobility on the performance of these distributed estimation algorithms.

This work was motivated by applications that have arisen in the eStadium sensor networks at Georgia Tech and Purdue. The research results described are from the PhD theses of Seema Bandyopadhyay and Xusheng Sun. The eStadium VIP teams at Georgia Tech, Morehouse College, and Purdue have carried out the design and implementation of the eStadium system.

Speaker Bio:

Edward J. Coyle received his B.S. degree from the University of Delaware in 1978, and the Ph.D. degree from Princeton University in 1982. From 1982 through 2007, he was a faculty member at Purdue University. Dr. Coyle joined Georgia Tech in 2008 and is currently the Arbutus Chair for the Integration of Research and Education, a Georgia Research Alliance Eminent Scholar, and director of the Arbutus Center for the Integration of Research and Education. The Vertically-Integrated Projects (VIP) Program is housed within the Arbutus Center and supports many large-scale, long-term R&D projects, including the eStadium project referred to in this talk.

His research interests include wireless and sensor networks, signal and image processing, and engineering education. He is a Fellow of the IEEE and a Distinguished Engineering Alumnus of the University of Delaware. He was a co-recipient of: the 2005 Bernard M Gordon Prize for Innovation in Engineering and Technology Education from the US National Academy of Engineering; the 1997 Chester F. Carlson Award for Innovation in Engineering Education; the IEEE ASSP Society’s 1986 Best Paper Award; and, the Myril B. Reed Best Paper Award from the 32nd Midwest Symposium on Circuits and Systems.


This will be a free event.


For questions or more information: A. Alfa, PhD 474-8789

Comments are closed.