Files in this item



application/pdf9503347.pdf (5MB)Restricted to U of Illinois
(no description provided)PDF


Title:Broadband microwave inverse scattering: Theory and experiment
Author(s):Weedon, William Harold
Doctoral Committee Chair(s):Chew, Weng Cho
Department / Program:Electrical and Computer Engineering
Discipline:Electrical Engineering
Degree Granting Institution:University of Illinois at Urbana-Champaign
Subject(s):Engineering, Electronics and Electrical
Abstract:This thesis presents both theoretical formulations and experimental methods for performing broadband time-domain inverse scattering. The inverse scattering problem is very important for a number of application areas including nondestructive evaluation, geophysical probing, medical imaging and military target identification. Emphasis is placed on the use of microwaves to probe the unknown object, although much of the theory presented applies to other types of waves such as acoustic and elastic waves.
Distorted-Born iterative method (DBIM) inverse scattering algorithms are presented for solving 2-D TM and TE problems. The TM algorithm is shown to be capable of accurately inverting objects with contrast as high as 10:1, but the TE algorithm breaks down when the object contrast exceeds 2:1 due to the buildup of polarization charges inside the object.
A local-shape-function (LSF) inverse scattering algorithm is presented for imaging very strong scattering objects such as metallic scatterers. The LSF algorithm has a higher resolution capability than the DBIM algorithm for reconstructing closely spaced metallic scatterers. The LSF algorithm also converges faster in the metallic scatterer case.
Broadband time-domain data are preferable to continuous-wave (CW) data at just a few discrete frequencies due to the higher information content inherent in a broadband pulse and the ability to use time gating to eliminate unwanted early-time and late-time arrival signals. Broadband time-domain data may be collected in a practical microwave measurement system using either a step-frequency radar approach or an impulse radar. There are advantages and disadvantages to using both data collection methods, and the choice of one technology over the other depends primarily on the application.
A prototype step-frequency radar system has been developed to demonstrate the capability of our inverse scattering algorithms with real experimental data. Reconstructions of both metallic and dielectric objects including metallic cylinders and plastic PVC pipes in air from experimental data are shown. A commercial monostatic impulse radar system is described and plans are discussed for building a rudimentary bistatic impulse radar.
A method is presented for implementing an efficient finite-difference time-domain (FDTD) electromagnetic scattering algorithm on a massively parallel supercomputer. The main challenge in designing an efficient algorithm is in the implementation of an absorbing boundary condition at the edge of the FDTD grid. Since the inverse scattering methods that we present here rely on the solution of forward scattering problems at each step in an iterative algorithm, the efficient FDTD algorithm allows us to solve very large inverse scattering problems quickly on a massively parallel supercomputer.
Issue Date:1994
Rights Information:Copyright 1994 Weedon, William Harold
Date Available in IDEALS:2011-05-07
Identifier in Online Catalog:AAI9503347
OCLC Identifier:(UMI)AAI9503347

This item appears in the following Collection(s)

Item Statistics