SYNTHETIC DATA PREDICTION AND VALIDATION TECHNIQUES FOR AUTOMATIC TARGET RECOGNITION(ATR)

Recent advances in full wave solutions to time-domain Maxwell's equations employing high order space and time, finite-element-like unstructured grid-based integration procedures, and parallel scalable code architectures, now make it very attractive and cost-effective to consider such solutions for a myraid of complex electromagnetic applications including radar cross section (RCS) scattering studies for full fighter targets, antenna radiation problems, and synthetic data prediction for automatic target recognition (ATR). While the current trend is to employ high frequency techniques for ATR studies, various aspects of electromagnetic physics, such as traveling waves, dispersive and other material characterization, interaction between electrically small and large components of an LO target (small fins on a long missile body), and low frequency (50 MHz to 2 GHz) modeling, require more exact solutions based on time-domain Maxwell's equations. The parallel, unstructured grid-based, time-domain CEM code, UPRCS, developed under the PACESETTER contract, will complement ongoing high frequency ATR applications.The key attributes of the time-domain UPRCS code are (1) pulse mode for broad band response, (2) complex geometry modeling using unstructured surface and volume gridding including material treatment, (3) highly scalable parallel code architecture, and (4) GUI for user friendly enviornment. While advances in the time-domain are being made in all fronts, the primary focus of this SBIR Phase I proposal is to develop the necessary procedures and validate the time-domain technology for ATR applications and extend the technology to X-band frequency range.