Geo-Statistical Analysis, Modeling and Simulation for Behavioral Motion Patterns

The Air Force has collected a large repository of Ground Moving Target Indicator (GMTI) data from a variety of sensors. This wealth of data offers an unprecedented opportunity to perform comprehensive statistical analyses of surface vehicle motions under a variety of conditions and to better understand the correlations between these vehicle patterns and operationally significant events. To date, a comprehensive analysis of this data utilizing modern time-series analyses and geo-statistical analysis based tools has not yet been performed against this data. In this research, we propose to perform this analysis with a specific goal of estimating the relationship between traffic models and various parameters of interest (including time of day, location, proximity to other events, etc). The insights gained into the statistical behaviors will then be applied as the basis for generative models to create a large-scale simulation capability suitable for producing scenarios several orders of magnitude larger in terms of both time and space than current capabilities.