A universal law of logarithm of the recurrence time (Q4430680)

Let \(T\) be a map of the interval \([0,1]\) into itself. For \(x\in T\) and \(n\in\mathbb N\) let \(K_n(x)\) be the \(n\)th recurrence time of \(x\): \[ K_n(x)=\min\{ j\geq 1: T^j (x)\in E_n(x)\}, \] where \(E_n(x)\) is one of the intervals \(E_{n,j}=[(j-1)/2^n,j/2^n)\), \(1\leq j\leq 2^n\) containing \(x\). For higher-dimensional cases the recurrence time is defined using subcubes instead of subintervals. For a class of \(T\) including Hénon mappings two conjectures are presented: A. If \(T\) is ergodic and has positive entropy then the sequence of averages of \((\log_2K_n)/n\) monotonically converges to the Hausdorff dimension as \(n\to\infty\). B. The values of \(K_nP_n\) are exponentially distributed as \(n\to\infty\), where \(P_n(x)\) is the measure of \(E_n(x)\). Rigorous computer simulations of various dynamical systems (the logistic map, the Gauss map, a toral automorphism, the Hénon map and others) are given to support these conjectures.