Dyadic approximation in the middle-third Cantor set (Q2107789)

\textit{K. Mahler} [Bull. Aust. Math. Soc. 29, 101--108 (1984; Zbl 0517.10001)] posed the following question: ``How close can irrational elements of Cantor's set be approximated by rational numbers (i) in Cantor's set, and (ii) by rational numbers not in Cantor's set?'' The present research deals with another aspect of Mahler's problem. That is, the main attention is given to the problem of how well points in the middle-third Cantor set can be approximated by dyadic rationals (i.e., rationals with denominators which are powers of 2). The authors note the following: ``We find that the behaviour when we consider dyadic approximation in the Cantor set is substantially different to considering triadic approximation in the Cantor set. In some sense, this difference in behaviour is a manifestation of Furstenberg's times 2 times 3 phenomenon from dynamical systems, which asserts that the base 2 and base 3 expansions of a number are not both structured.'' Proofs are explained, relations between known and present results are noted. The difference in behaviour between the dyadic and triadic approximation is considered. The relevance of counting dyadic rationals near the middle-third Cantor set is described. Certain attention is also given to the Hausdorff theory of Diophantine approximations on missing-digit Cantor sets and to the Cantor measure. It is noted that the Cantor measure is ``friendly'' measure in the parlance introduced by Kleinbock, Lindenstrauss, and Weiss. Also, ``this class of measures includes many well-known natural measures supported on fractals and manifolds, and has been a hugely successful medium for systematising the metric theory of Diophantine approximation''. A natural construction of the middle-third Cantor set is recalled. To count instances of dyadic rationals being close to triadic rationals in the Cantor set, Fourier analysis is used. Schwartz functions, bump functions, and tempered distributions are discussed. Finally, the authors give some discussion, in which certain open questions are explored.