Performance Metrics for Correlation and Tracking Algorithms

Military commanders require situational awareness to support real- time decision-making. To obtain information on possibly hostile entities in an area of interest, surveillance system, which receive information from sensors such as radars, intelligence, and other sources, are often used. One of the objectives of surveillance systems that track aircraft is the formation of a Single Integrated Air Picture (SIAP), that represents a coherent resolution of information. Correlation is the process by which sensor measurements and other information are combined to keep the SIAP up-to-date in real time. A correlator, which is the software implementation of a correlation methodology, must resolve ambiguities and conflicting information to provide an operationally useful synthesis of surveillance data. Possible ambiguities include missed tracks, extra tracks, or position and velocity errors. The methods developed in this thesis are designed for use in evaluating the performance of air surveillance systems, of which correlators are an integral part. Maneuvering or closely spaced aircraft pose difficult issues for air surveillance systems. These are addressed by the performance metrics. Using scripted test scenarios in a modeling and simulation environment, comparisons of correlators can be made using nonparametric statistical methods. An experiment constructed in this manner can be used to support acquisition decision making.Naval Postgraduate School Monterey CA is the author of 'Performance Metrics for Correlation and Tracking Algorithms', published 2001 under ISBN 9781423528777 and ISBN 1423528778.