“…To address the computational complexity, various linear [2,15,17,18] and nonlinear [19,20] sub-optimal estimators have been proposed to approximate the optimal estimator, but they are not good approximations in terms of the estimation criterion, since the criterion for these linear estimators (see Definition 5) differs from the criterion for the optimal estimator (see Definition 1). From [19,20], it is clear that the nonlinear estimators are not obtained according to either one of these two criteria above. As one of the most costefficient estimation schemes, the interacting multiple model (IMM) estimator is proven to be a good approximation for the optimal estimator, since it is able to obtain the estimate fairly close to the optimal one [21][22][23], which motivates us to apply the IMM estimator to UDP-like systems, and then explore its properties, especially the three aforementioned key aspects: stability, convergence, and performance.…”