Optimizing Track-to-Track Data Fusion for Variable Cases

Many Intelligence, Surveillance, and Reconnaissance (ISR) systems involve combining track data from multiple sources, leading to an improved track picture. A multitude of track fusion algorithms exist, some of which require significant computational resources and have difficulty running in real-time when thousands of tracks must be maintained. Ideally, we would like to evaluate the utility of the various track fusion algorithms for a particular system. An analysis of different track fusion algorithms based on real data from the ISR system in which they will be deployed would yield the most accurate assessment of track fusion performance; however, this requires data from real experiments, which is often difficult to obtain in any quantity. An alternative is to test and evaluate algorithms within a simulated environment that can model to some degree of realism the conditions under which the system will operate (and can vary those conditions to conduct thorough trade studies). On this effort, Toyon will develop a prototype ISR system evaluation tool that combines our scenario simulation application, tracking applications, track fusion application, and performance analysis tools into a software system for evaluating track fusion performance under varying conditions (and algorithm choices) for an ISR system such as the BAMS.