The \(p\)-adic diaphony of the Halton sequence (Q372728): Difference between revisions

Let \(\omega=(\mathbf{x}_n)\) be a sequence in \([0,1)^s\). Let \(\mathbb{P}\) denote the set of prime numbers. For \(\mathbf{p}=(p_1, \dots, p_s)\in \mathbb{P}^s\) the \(\mathbf{p}\)-adic diaphony \(F_N(\omega)\) of the sequence \(\omega\) was defined by \textit{P. Hellekalek} [Acta Arith. 145, No. 3, 273--284 (2010; Zbl 1243.11082)]. The worst-case integration error in a function space \(\mathcal{H}\) on \([0,1)^s\) with norm \(\|\cdot\|\) is given by \[ e(\mathcal{H},\omega)=\sup_{f\in\mathcal{H}, \|f\|\leq 1}\left|\int_{[0,1]^s}f(\mathbf{x})\,d\mathbf{x}-\frac{1}{N}\sum_{n=0}^{N-1}f(\mathbf{x}_n)\right|. \] In this paper it is shown that the worst-case integration error in a certain reproducing kernel Hilbert space \(\mathcal{H}_{\mathbf{p},s}\) and the \(\mathbf{p}\)-adic diaphony \(F_N(\omega)\) of the sequence \(\omega\) are related by \(e(\mathcal{H}_{\mathbf{p},s},\omega)=\sqrt{\sigma_{\mathbf{p}}-1}F_N(\omega)\) where \(\sigma_{\mathbf{p}}=\prod_{i=1}^s(p_i+1)\). The main result of this paper is that the \(\mathbf{p}\)-adic diaphony \(F_N(\omega)\) of the Halton sequence \(\omega\) in pairwise different prime base \(\mathbf{p}=(p_1, \dots, p_s)\) satisfies \(F_N(\omega)=O((\log N)^{s/2}/N)\).