ECE Graduate Seminar
<p> The Electrical and Computer Engineering Department cordially invites you to attend the following seminar.</p> <p> </p> <p> </p> <p> Adaptive Information Fusion and</p> <p> Sensor Management in Networked Systems</p> <p> by</p> <p> Dr. Ruixin Niu</p> <p> Department of Electrical and Computer Engr.</p> <p> Virginia Commonwealth University</p> <p> Richmond, VA</p> <p> Friday, October 12, 2012</p> <p> 3:00 p.m. KH 224</p> <p> Host: Dr. Dimitrie Popescu</p> <p> </p> <p> This talk will cover two aspects of Dr. Ruixin Niu’s current research on information fusion and sensor management in networked systems. In the first half of the talk, an inference-driven adaptive sensor management framework will be presented, which is based on the new recursive conditional posterior Cramer-Rao lower bound (C-PCRLB) on the mean squared error (MSE), derived for Bayesian nonlinear filtering problems. In comparison with the state of the art sensor management approaches based on entropy or mutual information, the C-PCRLB has significantly reduced complexity and is directly related to the estimation performance in terms of the MSE. Examples, such as sensor selection and bandwidth management in WSNs, will be presented to illustrate the advantages of the proposed approach.</p> <p> </p> <p> The second part presents a new localization and tracking approach in a widely separated noncoherent multiple-input multiple-output (MIMO) radar systems. In the proposed MIMO radar system, the signals received at distributed receivers are processed jointly and the matched filter outputs are directly used for target tracking in a track-before-detect framework. It is demonstrated that the noncoherent MIMO radar with widely spaced transmit and receive elements can provide tracking accuracies that are significantly higher than that of a monostatic phased array radar with high range and azimuth resolutions. </p>
Posted By: Linda Marshall
Date: Mon Oct 08 13:29:10 EDT 2012