Multilayered Wall Characteristic Extraction for Through Wall Radar Systems

The primary objective of this Phase I project is to develop radar-based techniques and algorithms to accurately estimate the parameters of walls, such as dielectric constant, conductivity (or loss-tangent) and thickness, as well as to devise methods that can characterize inhomogeneous walls by detecting presence of foiled-back insulation, reinforced wires/studs, water pipes, and enclosed air-gaps, e.g., in hollow concrete (cinder-block) walls. The wall parameter estimation can be performed either in time-domain (suitable for modulated pulse or impulse radar operation) or in frequency-domain (suitable for CW or step-frequency radar operation). To achieve this we propose to apply and compare the accuracy and robustness of several alternative techniques including time-domain Reflectometry, Least-Squares method, Singularity Expansion Method (SEM), Evolutionary based optimization and Ellipsometry. In addition, we will investigate the following techniques for characterization of inhomogeneous walls: i) a technique based on the frequency signature of calibrated wall"s reflected signal to detect the presence of conductor a back-plated wall, ii) a hybrid of"effective"parameter estimation and subsurface imaging to characterize and image existence and location of re-bars, air-gaps and/or other wall"s interior structures, and iii) use of cross-polar scattered fields or other polarization-based radar techniques for detection of reinforced wires or water pipes and their orientations.