Conservativeness and uniqueness of invariant measures related to non-symmetric divergence type operators

TL;DR

This paper establishes criteria for the conservativeness and uniqueness of invariant measures for divergence type operators, ensuring well-behaved stochastic processes without requiring growth conditions on certain derivatives.

Contribution

It introduces new conservativeness criteria based on $L^1$-uniqueness and explores conditions for invariant measure uniqueness without growth restrictions on anti-symmetric matrix derivatives.

Findings

Criteria for conservativeness derived from $L^1$-uniqueness.

Conditions ensuring invariant measure uniqueness beyond recurrence.

Applicability to existence and uniqueness of strong solutions to associated SDEs.

Abstract

We present conservativeness criteria for sub-Markovian semigroups generated by divergence type operators with specified infinitesimally invariant measures. The conservativeness criteria in this article are derived by $L^1$-uniqueness and imply that a given infinitesimally invariant measure becomes an invariant measure. We explore further conditions on the coefficients of the partial differential operators that ensure the uniqueness of the invariant measure beyond the case where the corresponding semigroups are recurrent. A main observation is that for conservativeness and uniqueness of invariant measures in this article, no growth conditions are required for the partial derivatives related to the anti-symmetric matrix of functions $C=(c_{ij})_{1 \leq i,j \leq d}$ that determine a part of the drift coefficient. As stochastic counterparts, our results can be applied to show not only the existence of a pathwise unique and strong solution up to infinity to a corresponding It\^{o}-SDE, but also the existence and uniqueness of invariant measures for the family of strong solutions.