On the average number of unitary factors of finite abelian groups (Q1959781)

From MaRDI portal
Revision as of 19:58, 29 July 2023 by Importer (talk | contribs) (‎Created a new Item)
(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)
scientific article
Language Label Description Also known as
English
On the average number of unitary factors of finite abelian groups
scientific article

    Statements

    On the average number of unitary factors of finite abelian groups (English)
    0 references
    0 references
    12 October 2010
    0 references
    \textit{E. Cohen} [Acta Arith. 6, 159--173 (1960; Zbl 0113.25305)] considered the monoid \(X\) of isomorphism classes of finite abelian groups with multiplication defined by the direct product. He called \(A\in X\) a unitary factor of \(G\in X\) if, for some \(B\in X\), we have \(AB=G\), and \(A\) and \(B\) are relatively prime, i.e. have no non-trivial common divisor in \(X\). Let \(t(G)\) denote the number of such factors. For example \(t(C(2)\times C(2))=3\) as we have one non-trivial and two trivial decompositions. Let \[ T^*(x) = \sum_{G\in X, |G| \leq x} t(G). \] Cohen has shown an estimate of the form \[ T^*(x) = c_1 x \log x + c_2 x + O(\sqrt{x} \log x). \] \textit{E. Krätzel} [Acta Arith. 51, No. 4, 369--379 (1988; Zbl 0633.10044)] has refined this to \[ T^*(x) = c_1 x \log x + c_2 x + c_3 \sqrt{x} + \Delta^*(x) \] for some \(\Delta^*(x)=O(x^{11/29} \log^2 x)\). For positive integers \(k,n\) let \(t_k(n)\) be such that \(\sum_{n=1}^\infty t_k(n)n^{-s}=\prod_{j=k}^\infty\zeta^{1+\varepsilon(j)}(js)\), where \(\zeta(s)\) is Riemann's zeta function and \(\varepsilon(j)\) equals \(1\) for odd \(j\) and \(0\) for even \(j\). Krätzel's approach was to express \(T^*(x)\) in terms of \(t_3(n)\) and the function \[ D(1,1,2;x) = \sum_{n_1 n_2 n_3^2 \leq 1} 1. \] The function \(D(1,1,2;x)\) is a sum of a simple main term, a ''middle term'' involving some exponential sums, and a remainder of the order \(O(x^{1/4})\). Krätzel's proof shows that if the middle term is \(O(x^{\theta+\varepsilon})\) for some \(\theta\geq 1/4\), then so is \(\Delta^*(x)\). The present author has given an improved upper bound for the middle term obtaining \(\theta = 63/178 < 0.35394\).
    0 references
    exponential sums
    0 references
    abelian groups
    0 references

    Identifiers