Ergodicity of supercritical SDEs driven by $\alpha$-stable processes and heavy-tailed sampling

TL;DR

This paper proves exponential ergodicity for a class of supercritical SDEs driven by $ ext{alpha}$-stable processes, using heat kernel estimates, and introduces a new heavy-tailed sampling method.

Contribution

It establishes $V$-uniform exponential ergodicity for supercritical SDEs driven by $ ext{alpha}$-stable processes under new conditions, extending understanding of ergodic behavior in heavy-tailed settings.

Findings

Proved $V$-uniform exponential ergodicity for the SDEs.

Utilized heat kernel estimates to establish strong Feller property and irreducibility.

Introduced a novel heavy-tailed sampling scheme.

Abstract

Let $\alpha\in(0,2)$ and $d\in\mathbb{N}$. Consider the following stochastic differential equation (SDE) driven by $\alpha$-stable process in $\mathbb{R}^d$: $$ dX_t=b(X_t)dt+\sigma(X_{t-})d L^{\alpha}_t, \quad X_0=x\in\mathbb{R}^d, $$ where $b:\mathbb{R}^d\to\mathbb{R}^d$ and $\sigma:\mathbb{R}^d\to\mathbb{R}^d\otimes\mathbb{R}^d$ are locally $\gamma$-H\"older continuous with $\gamma\in(0\vee(1-\alpha)^+,1]$, $L^\alpha_t$ is a $d$-dimensional rotationally invariant $\alpha$-stable process. Under some dissipative and non-degenerate assumptions on $b,\sigma$, we show the $V$-uniformly exponential ergodicity for the semigroup $P_t$ associated with $(X_t(x),t\geq 0)$. Our proofs are mainly based on the heat kernel estimates recently established in \cite{MZ20} through showing the strong Feller property and the irreducibility of $P_t$. It is interesting that when $\alpha$ goes to zero, the diffusion coefficient $\sigma$ can grow faster than drift $b$. As applications, we put forward a new heavy-tailed sampling scheme.