ASSET: A simulation test bed for data association algorithms

Data association is an information processing procedure typically associated with surveillance and monitoring activities. Because of the importance of data association in a variety of areas from military operations to meteorology and medicine, there are a number of algorithms that purport to effectively automate it. Evaluation of these algorithms on realistic problems is difficult because until recently there was no accepted measure for algorithm performance. This paper describes ASSET, a test bed for data association algorithms, that evaluates the effectiveness of these algorithms from a user’s perspective. The test bed produces performance measures for the algorithms, which allow for easy comparisons over a range of design parameters. Applications of the test bed to the testing of current surveillance algorithms are also described.