A description is given of a global modeling approach developed for the multitarget tracking (MTT) problem. From a global modeling of this problem, both data association (DA) and maneuver estimation problems can be simultaneously solved using system identification techniques. With this approach, previously developed single-target tracking/acceleration estimation techniques can be directly used for the MTT problem. Both DA and MDE (maneuver detection and estimation) are viewed as simultaneous problems where possible interactions among the targets can be taken into consideration. The global modeling translates this problem to a system identification problem for systems with time-varying parameters where the parameters can take only a finite number of values at each time. This becomes an adaptive (multiple model) Kalman filtering (MMKF) problem whose solution can be computed. The computational load associated with the full solution of MMKF can then be reduced utilizing many types of approximate (suboptimal) solutions developed in the literature. >