Methodologies for Accurate Assessment of Target Characteristics

Accurate knowledge of a target's radar signature is a critical issue in assessing the performance of the Ballistic Missile Defense (BMDS) system, and digital signature models continue are used extensively in simulation of BMDS elements. Though existing predictive signature codes are fast, they are limited in their ability to model non-specular scattering effects and dielectric materials. The need therefore exists for software codes that will correctly model these, while producing data quickly at the needed sampling rates. We propose to develop a predictive signature code, using the Method of Moments (MOM) and the Fast Multipole Method (FMM), that will address these issues. A key element of this new code will be its use of hardware and software acceleration strategies, leveraging threads, CPU vector instructions, and GPU computing. In Phase I, we will develop a proof-of-concept prototype software, and verify its validity using electromagnetic theory and measurements. The prototype software will be further developed and refined during the Phase II effort.