“…The three basic functions -data association, track maintenance, and filtering and prediction -are now examined. Then, the match can be taken to be a probability, and written: (5) where M is the number of dimensions in the Where space, and x 11 and A 11 are, respectively, the Kalman filtering (Bar-Shalom and Li, 1993;Blackman, 1986) is employed to predict the next position Xiz, and covariance matrix A 11 of each track h. Recall from Section 3 that the usual clustering is bypassed when b corresponds to a PDW that has been assigned a previouslyestablished track through TOA deinterleaving. In this case, b retains its track, and does not perform data association, nor track maintenance.…”