Quantitative recurrence problem on some Bedford-McMullen carpets
Yu-Liang Wu, Na Yuan
TL;DR
This work computes the Hausdorff dimension of quantitative recurrent sets for the canonical endomorphism on Bedford-McMullen carpets in the regime where the carpet's Hausdorff and box dimensions coincide. The authors develop a probabilistic, self-affine framework, constructing a piecewise Bernoulli measure to capture the local dimension and employing a Falconer-type argument to obtain a sharp lower bound, then complementing it with a delicate covering argument to yield the matching upper bound. The main result expresses dim_H W_gamma(K,T,psi) as min{t1,t2}, with explicit formulas for t1 and t2 depending on scale parameters tau1, tau2, entropy terms H1(p), H2(p), and the column-count data M,N, plus a gamma-dependent reduction to simpler cases when gamma=1. The paper also demonstrates concrete applications to diagonal endomorphisms and Cantor-set products, illustrating how the general formula specializes to familiar non-conformal settings and highlighting the role of self-affine geometry in shrinking-target-type problems.
Abstract
In this paper, we study the Hausdorff dimension of the quantitative recurrent set of the canonical endomorphism on the Bedford-McMullen carpets whose Hausdorff dimension and box dimension are equal.
