A distributed multirate IMM algorithm for multiplatform tracking

A distributed multirate interacting multiple model (DMRIMM) algorithm for m ulti-platform tracking is proposed. Under the assumption that each model op erates at an update rate proportional to the model's assumed dynamics, a wa velet transform is used to generate equivalent lent multirate measurements and a multirate constant-velocity (MRCV) model is derived. Thus, a multirat e interacting multiple model (MRIMM) algorithm using an MRCV model and mult irate measurements is obtained. Based on the multirate IMM, its distributed version DMRIMM algorithm for multiplatform tracking is proposed. The MRIMM algorithm is first employed to perform each local/platform estimation. A g lobal filter is then constructed to perform a fusion of MRIMM estimations. The advantages of low computation loads and performance improvement are dem onstrated through Monte Carlo simulations. (C) 2000 Elsevier Science Ltd. A ll rights reserved.