Approaches to Multisensor Data Fusion

s part of an Office of Naval Research–funded science and technology development task, APL is developing an identification (ID) sensor data fusion testbed. The testbed is driven by an APL-modified version of the Joint Composite Tracking Network pilot benchmark called the Composite Combat ID Analysis Testbed (CAT). The CAT provides accurate tracking for realistic scenarios involving multiple targets and netted radar and ID sensors. Track state outputs from the CAT include feature information from electronic support measures and noncooperating target recognition sensors. These data are combined to improve the confidence of aircraft-type declarations using both Bayesian and evidential reasoning–based algorithms. We use measure theoretic methods to describe the relationship between Bayesian theory and the Dempster-Shafer evidential reasoning theory.