High Performance Computing (HPC) for Real-Time Course of Action (COA) Analysis

摘要

The focus of this research project was to develop and demonstrate proof of concept technologies to assist decision makers in assessing friendly effects-based Course of Actions (COAs) against an intelligent adversary in an operational-level simulation environment, faster than real-time. The R&D activities included multiple components: a simulation test bed; a scalable, flexible simulation framework; automated scenario generation techniques with dynamic update; intelligent adversarial behavior modeling; effects- based/attrition-based behavior modeling; and real-time analysis technology for comparing and grading effectiveness of alternative simulations. The following highlights the in-house accomplishments: developed an effects-based center of gravity (COG) modeling capability and an in-house COA simulation test bed capable of simulating direct, indirect, cumulative, cascading and recovery events. The following complimentary capabilities were also developed under extramural research projects: semiautomated scenario generation that produced COAs in minutes/hours vs days, simulation cloning on HPCs for parallel COA evaluation, dynamic COA analysis incorporating an unscripted adversary, and COA simulation comparisons produced in seconds vs hours.