Measures of Effectiveness for the Distributed Data Fusion Problem

摘要

This report surveys measures of effectiveness applicable to data fusion systemsaddressing Level 1 of the taxonomy of data fusion processes identified by the Joint Directors of Labs Data Fusion Subpanel, particularly as these measures apply to the detection, association, and localization subproblems, with less attention paid to the classification subproblem. The measures are presented starting with those applicable to the simplest fusion systems and ending with the most complex type, i.e., the multiple-target, multiple-sensor, multiple-hypothesis, distributed data fusion system. The applicability of each and their shortcomings are discussed, with the perspective oriented toward applications involving Navy antisubmarine warfare.