A GENETIC ALGORITHM-BASED MULTIDIMENSIONAL DATA ASSOCIATION ALGORITHMFOR MULTI-SENSOR-MULTI-TARGET TRACKING

The central problem in multitarget-multisensor tracking is the data as sociation problem of partitioning the observations into tracks and fal se alarms so that an accurate estimate of true tracks can be found. Th e data association problem is formed as an N-dimensional (N-D) assignm ent problem, which is a state-of-the-art method and is NP-hard for N g reater than or equal to 3 sensor scans. This paper proposes a new gene tic algorithm for solving the above problem which is typically encount ered in the application of target tracking. The data association capac ities of the genetic algorithm have been studied in different environm ents, and the results are presented.