Information Fusion and Prediction for Space Situational Awareness
Small Business Information
DECISIVE ANALYTICS Corporation
1235 South Clark Street, Suite 400, Arlington, VA, 22202
AbstractUnder this SBIR, the Decisive Analytics Corporation (DAC) team proposes a novel solution to creating a clear and timely space asset threat picture from available intelligence data by recognizing and anticipating hostile space activities. The DAC solution is a supervised learning approach that mines the temporal structure of complex space events from historical data and uses these temporal relations as features in the characterization and anticipation of future space events. The solution involves two processing loops: 1. A training loop in which important discriminating relationships fingerprints comprising different hostile and unintentional space activities are learned from historical data; and 2. An on-line event recognition loop that discovers patterns and composite relationships in the available data and uses these mined relations to characterize the hostility of events in the evolving space scene and to anticipate events that are likely to occur in the future. The proposed approach builds on DACs experience in designing, implementing and delivering advanced data mining and association algorithms to the DoD community. These technologies, combined with the DAC teams experience in planning, coordinating, managing and executing worldwide Space Situational Awareness operations to detect, track, identify, catalog and protect international space systems will result in algorithms suitable for prototyping in Phase I and Phase II and deployment to the US Air Force in Phase III of this SBIR. BENEFIT: The anticipated benefit of the proposed solution is an improvement in the ability to utilize all relevant threat intelligence information to detect intentional space threats, unambiguously classify anomalous behavior as environmental, man-made unintentional or hostile, and anticipate hostile space actions. This will include a logical and easily interpretable graph of the information that led to a particular classification determination i.e. the chain of evidence needed to infer the hostility of a space action. The approach is also applicable to the characterization of threats in the areas of missile defense and to the classification of complex events (activities) in video.
* information listed above is at the time of submission.