A lower bound on the integral of the nonlinear filtering error (Q1080828)

Since in most cases easily computable nonlinear filters are not available, it is of great importance, for the sake of deriving good suboptimal filters, to obtain tight bounds on the nonlinear filtering error. This paper presents a lower bound of the nonlinear filtering error, in the case of a one-dimensional unknown process observed linearly (or possibly through a one-to-one mapping) with additive white noise.