A pointwise ergodic theorem along return times of rapidly mixing systems

Abstract

We introduce a new class of sparse sequences that are ergodic and pointwise universally $L^2$-good for ergodic averages. That is, sequences along which the ergodic averages converge almost surely to the projection to invariant functions. These sequences are generated randomly as return or hitting times in systems exhibiting a rapid correlation decay. This can be seen as a natural variant of Bourgain's Return Times Theorem. As an example, we obtain that for any $a\in (0,1/2)$, the sequence $\left\{n\in \mathbb{N}:2^{n}y\mathrm{mod}1\in (0,n^{-a})\right\}$ is ergodic and pointwise universally $L^2$-good for Lebesgue almost every $y\in [0,1]$. Our approach builds on techniques developed by Frantzikinakis, Lesigne, and Wierdl in their study of sequences generated by independent random variables, which we adapt to the non-independent case.

15 pages. Suggestions and comments from the referees have been incorporated, to appear in Bulletin of the London Mathematical Society