Automated Adversarial Course of Action Model Generation and Reasoning for Satellite Protection (commercial/military)

ABSTRACT: We have shown in Phase I the technology provides unique benefits in scenario metrics, what-if analysis, and action/response assessment. We have also demonstrated a Phase I prototype that validates application in the relevant domain and the remaining work required to develop a transitionable adversarial threat assessment tool in Phase II. We propose to design, develop, and deliver a ground breaking tool that will significantly aide analyst work in predicting and combating such threats. The Space Assessment Simulator (SAS) will include complex models of relevant naturally occurring events, as well as models of potential adversary resources including both kinetic and cyber capabilities. SAS will apply powerful game-theoretic algorithms to deduce a wide range of possible threats, ranging from most likely to most dangerous. It will include a sophisticated timing model to capture possibilities of adversaries launching attacks coordinated with natural events to disguise or confuse interpretation of their actions. These threat trees will be passed to the analysts for use in developing appropriate countermeasures. BENEFIT: The Space Assessment Simulator contains a library of possible actions, goals, and domains that are potentially applicable to either the defender of the space assets, or an adversary. Player models are then developed that contain a subset of these moves available to each player. Potential interactive natural events are also stored. The Space Assessment Engine can then examine millions of possibilities, and output rank ordered threats.