Filtering linear systems with large time-varying measurement delays

In this paper we consider the estimation problem for linear stochastic systems affected by multiple known and time-varying delays on all the output signals. Based on a modification of a previous proposal we prove for the first time the result that a filter based on simple eigenvalue assignment of the closed-loop error system may achieve uniform performance, with respect to the delay bound and estimation variance, in presence of both constant and time-varying delays that are differentiable. A new and simple demonstration technique provides non conservative delay bounds for time-varying delays. A cascaded version of the filter can cope with arbitrarily large delays.