Gary Haith (Referentia Systems, Inc.), Stephen C. Upton (Referentia Systems Incorporated)
Keywords: Modeling, Analysis, Simulations
Abstract:
Satellites and other spacecraft are key assets and critical vulnerabilities in our communications, surveillance and defense infrastructure. Despite their strategic importance, there are significant gaps in our real-time knowledge of satellite security. One reason is the lack of infrastructure and applications to filter and process the overwhelming amounts of relevant data. Some efforts are addressing this challenge by fusing the data gathered from ground, air and space based sensors to detect and categorize anomalous situations. The aim is to provide decision support for Space Situational Awareness (SSA) and Defensive Counterspace (DCS). Most results have not yielded estimates of impact and cost of a given situation or suggested courses of action (level 3 data fusion). This paper describes an effort to provide high level data fusion for SSA/DCS though two complementary thrusts: threat scenario simulation with Automatic Red Teaming (ART), and historical data warehousing and mining. ART uses stochastic search algorithms (e.g., evolutionary algorithms) to evolve strategies in agent based simulations. ART provides techniques to formally specify anomalous condition scenarios envisioned by subject matter experts and to explore alternative scenarios. The simulation data can then support impact estimates and course of action evaluations. The data mining thrust has focused on finding correlations between subsystems anomalies on MightySat II and publicly available space weather data. This paper describes the ART approach, some potential correlations discovered between satellite subsystem anomalies and space weather events, and future work planned on the project.
Date of Conference: September 12-15, 2007
Track: Modeling, Analysis and Simulations