On a conjecture of Dekking: the sum of digits of even numbers

From MaRDI portal
Publication:2454441



Abstract: Let qgeq2 and denote by sq the sum-of-digits function in base q. For j=0,1,...,q−1 consider # {0 le n < N : ;;s_q(2n) equiv j pmod q }. In 1983, F. M. Dekking conjectured that this quantity is greater than N/q and, respectively, less than N/q for infinitely many N, thereby claiming an absence of a drift (or Newman) phenomenon. In this paper we prove his conjecture.


In the present paper the authors consider a conjecture of \textit{F. M. Dekking} [Sémin. Théor. Nombres, Univ. Bordeaux I 1982--1983, Exp. No. 32, 12 p. (1983; Zbl 0529.10047)] involving the sum of digits function of even numbers. The used methods are discrete Fourier analysis and generating functions. Let \(q\geq2\) an integer and \(s_q\) the sum of digits function in base \(q\). Then for fixed integers \(q\), \(d\), \(m\geq2\) and \(a\) and \(j\) the authors consider the error term \(g(N)\) in \[ \#\{0\leq n\leq N: n\equiv a\bmod d, s_q(n)\equiv j\bmod m\} =\frac{N}{dm}+g(N). \] In particular, Dekking conjectured the following drifting (or Newman-type) phenomenon: for \[ (q,a,d,j,m)\equiv(q,0,2,j,q) \] we have that \(g(N)<0\) and \(g(N)>0\) infinitely often. In the present paper the authors consider the more general case \[ (q,a,d,j,m)\equiv(q,a,d,j,q). \] Then they show that if \(d\mid q\) then \(g(N)=\mathcal{O}(1)\) and \(g(N)\) changes its sign infinitely often and if \(d\mid q-1\) then \(g(N)\) can take arbitrarily large positive as well as arbitrarily large negative values as \(N\to\infty\). For \(d=2\) this proves Dekking's conjecture and covers all bases \(q\geq2\).











This page was built for publication: On a conjecture of Dekking: the sum of digits of even numbers

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q2454441)