IMM Estimator for Maneuvering Target Tracking in GSM Networks

Global system for mobile communications (GSM) networks can provide position information. For the application of tracking ground target, the target should be regarded as maneuvering target since the state to be estimated may change dramatically due to the restriction of terrain, road, traffic, and so on. It is well known that multiple model (MM) estimation is a powerful approach for such a hybrid estimation problem. In this paper, we consider this target motion uncertainty problem in GSM environment. Interacting MM (IMM) estimator is analyzed and the estimation results of three different simulated scenarios are evaluated by the corresponding posterior Cramer-Rao lower bound (PCRB) and compared with single model method of using one nearly constant velocity (CV) model. The comparisons show that IMM estimator reduce the estimation errors during the non-maneuvering period.