A new approach to the existence of invariant measures for Markovian semigroups (Q2320390): Difference between revisions

The purpose of the paper is to give a new two-step approach to prove the existence of finite invariant measures for a given Markovian semigroup. First, it introduces a convenient auxiliary measure and then gives conditions equivalent to the existence of an invariant measure which is absolutely continuous with respect to it. The main results are entirely measure-theoretic and do not involve irreducibility properties of the semigroup. Several applications are considered. Section 3.1 contains a unifying generalization of various versions of Harris' ergodic theorem to a more general one. Corollary 3.8 gives an answer to an open question of \textit{R. L. Tweedie} [Stochastic Processes Appl. 92, No. 2, 345--354 (2001; Zbl 1047.60068)]. Section 3.2 considers a nonlinear stochastic partial differential equations on a Gelfand triple. It is shown that under a Wang's Harnack-type inequality, the strict coercivity condition is sufficient to guarantee the existence of a unique invariant probability measure for the solution. The authors also consider a perturbation of a Markov kernel satisfying a combined Harnack-Lyapunov condition, for which the result of R.L. Tweedie cannot be used, but their two-step approach works easily. They also discuss the applicability of Harris' result to this kind of perturbation.