An infinite interval version of the \(\alpha \)-Kakutani equidistribution problem (Q6561667)

For a fixed \(0<\alpha < 1\), the \(\alpha\)-Kakutani scheme is a sequence of partitions which is defined inductively: The first partition consists of the trivial partition, i.e., \(P_0=\{ [0,1] \}\) and \(P_{n+1}\) is obtained from \(P_n\) by subdividing each interval of maximal length into two smaller intervals in the ratio \(\alpha : 1-\alpha\).\N\NDenoting the set of endpoints of the intervals in the \(n\)-th stage of this process with \(E_n\), then \textit{S. Kakutani} [Lect. Notes Math. 541, 369--375 (1976; Zbl 0363.60023)] proved that for all \(\alpha \in (0,1)\) the set \(E_n\) becomes uniformly distributed as \(n \rightarrow \infty\).\N\N\textit{A. Volčič} [Ann. Mat. Pura Appl. (4) 190, No. 1, 45--54 (2011; Zbl 1219.11113)] generalised this process by splitting the intervals of maximal length according to a fixed finite partition while the authors go one step further and insert an infinite partition into each interval of maximal length. In this case, the definition of the set \(E_n\) has to be modified and consists of the endpoints of those intervals which have been split up to the \((n+1)\)st stage. Using this modified definitions the authors derive a generalisation of Kakutani's Theorem (Theorem 1). Furthermore, the authors derive discrepancy estimates (Theorem 2, Theorem 3).