Print

Print


Notice and Invitation

Oral Defense of Doctoral Dissertation
The Volgenau School of Engineering, George Mason University


Hossein Roufarshbaf

Bachelor of Science, Isfahan University of Technology, 1999
Master of Science, Amirkabir University of Technology, 2002


A Tree Search Approach to Detection and Estimation with Application
to
Communications and Tracking


Wednesday, July 20, 2011, 1:00pm-3:00pm
Nguyen Engineering Bldg., Room 3507


All are invited to attend.


Committee

Jill K. Nelson, Chair
Andre Manitius
Kuo Chu Chang
Gerald Cook


Abstract
A novel approach to detection and estimation problems using tree search techniques is presented. The problem is framed as a generalized sequential detection problem in
which every possible sequence of system states is mapped to a path through the search tree. The stack algorithm and the M-algorithm that are originally used in
decoding of the convolutional codes are implemented to reduce the computational complexity of the tree search technique. The proposed tree search technique can be
viewed as approximating the full Bayesian inference approach by computing the posterior distribution only in regions in which it has significant mass. Unlike
approaches that build on Kalman filtering techniques, the proposed stack-based tracker suffers no performance loss in the presence of nonlinear and/or non-
Gaussian system state space model. The proposed algorithm is successfully applied to blind channel equalization, modulation classification and target trackingproblems.


A copy of this doctoral dissertation is on reserve at the Johnson Center Library.