Signature Prediction and Uncertainty Analysis for Recognition Applications

The goal of signature prediction is to provide electromagnetics modeling to support detection and identification technologies for a wide range of target configurations and sensor characteristics. Signature prediction technologies have been extensively developed by AFRL, leading to a suite of integrated codes used for generating radar signatures of both air and ground targets. An important input to the codes is a CAD file of the target geometry. When selected from the CAD library, it is desired that the model contains reliable materials information that doesn’t cause undue uncertainty in the resulting signature predictions. The material information is typically measured using a handheld RF tool. The existing tool has no bandwidth for characterizing materials across frequencies, provides limited RF material information, and has no data storage capability. In this SBIR, BerrieHill Research shall demonstrate an advanced RF material measurement tool having the ability to fully characterize material surfaces and computerize the data collection. Such a tool will lead directly to more accurate material information that can easily be incorporated into the target CAD file library, resulting in better signature predictions that don’t suffer uncertainty due to unreliable or incomplete materials information.