Pretentiously detecting power cancellation (Q2842340)

From MaRDI portal

!

This is the item page for this Wikibase entity, intended for internal use and editing purposes. Please use the normal view instead:

scientific article; zbMATH DE number 6198185
Language Label Description Also known as
default for all languages
No label defined
    English
    Pretentiously detecting power cancellation
    scientific article; zbMATH DE number 6198185

      Statements

      Pretentiously detecting power cancellation (English)
      0 references
      0 references
      0 references
      13 August 2013
      0 references
      multiplicative functions
      0 references
      pretentiousness
      0 references
      power cancellation
      0 references
      Let \(f\), \(g\) be multiplicative functions taking values in the complex unit disc, and define NEWLINE\[NEWLINE\mathbb D(f,g)^2:= \sum_p{1-\text{Re}(f(p)\overline g(p))\over p}.NEWLINE\]NEWLINE When this quantity is finite, Granville and Soundararajan called the functions \(f(n)\) and \(g(n)\) pretentious to each other, and they have written a series of papers on this topic; see, for example, the report by \textit{A. Granville} [J. Théor. Nombres Bordx. 21, No. 1, 159--173 (2009; Zbl 1236.11086)]. The aim of the present paper is to extend the idea of pretentiousness in two different ways in order to preserve power cancellation. NEWLINENEWLINEFor any multiplicative functions \(f(n)\), \(g(n)\) define the multiplicative function \(h(n)\) by \(g(n)= (f* h)(n)\), where the right-side represents the Dirichlet convolution. For \(\beta>0\) let NEWLINE\[NEWLINE\begin{aligned} H_\beta(f,g) &:= \sum_p \sum^\infty_{k=1} {|h(p^k)|\over p^{k\beta}},\\ \widehat{\mathbb D}_\beta(f,g) &:= \sum_p \sum^\infty_{j=1} {|g(p^j)-f(p^j)|\over p^{j\beta}}.\end{aligned}NEWLINE\]NEWLINE The authors explore the interrelationship between the finiteness of \(H_\beta(f,g)\) and \(\widehat{\mathbb D}_\beta(f,g)\) for some or all \(\beta> 0\) under various assumptions. They use these functions to determine conditions under which \(S_f(x):=\sum_{n\leq x} f(n)\ll x^\alpha\) \((\alpha>0)\) implies that \(S_g(x)\ll x^{\max(\alpha,\beta)}\) or just \(S_g(x)\ll x^\alpha\). Altogether the authors establish six theorems, and their proofs involve some manipulations of multiple sums and other techniques from linear algebra and complex analysis.
      0 references

      Identifiers