I. Introduction
Filtering or state estimation problem is one of the mostly investigated topics in signal processing and control communities. Ever since the pioneering work of Luenberger in 1960s [27], there has been an ongoing research interest in the filtering or state estimation problems for various systems [23]. For example, in [32], the optimal filtering problem has been investigated for the linear time-invariant (LTI) system with unknown covariances of process noises and measurement noises. In [22], a so-called Gaussian particle filter has been developed for nonlinear systems through approximating the posterior distributions by single Gaussian models. In [15], the robust state estimation issue has been studied for uncertain systems with limited communication capacity based on a parameter-dependent (PD) approach. The moving horizon estimation problem has been studied in [48] and [49] for networked systems with unknown inputs and communication constraints, respectively.