The inverse problem for representation functions of additive bases

Density, gaps, topology (11B05) Additive bases, including sumsets (11B13) Representation functions (11B34) Inverse problems of additive number theory, including sumsets (11P70)

Review text

A subset \({\mathcal A}\) of \(\mathbb Z\) is an asymptotic basis of order 2 if all but finitely many integers can be written in the form \(a+a'\) with \(a,a'\in{\mathcal A}\). If \(r_{\mathcal A}(n)\) is the representation function of \({\mathcal A}\), that is the number of ways to write an integer \(n\) as \(a+a'\) with \(a,a'\in{\mathcal A}\) and \(a\leq a'\), then \({\mathcal A}\) is an asymptotic basis if and only if the set \(r_{\mathcal A}^{-1}(0)\) is finite. Previously, the author in [Unique representation bases for the integers, Acta Arith. 108, No. 1, 1--8 (2003; Zbl 1154.11306)] had proved that thin asymptotic bases with a prescribed representation function did exist (see the reference for a precise result). In the paper under review, the author is interested in an existence result of thick asymptotic bases with a prescribed representation function. He proves that, for every function \(f:\mathbb Z\to\mathbb N\cup\{\infty\}\) such that the set \(f^{-1}(0)\) is finite, there exist asymptotic bases \({\mathcal A}\) of order 2, such that \(r_{\mathcal A}=f\) and \(|{\mathcal A}\cap[-x,x]|\gg x^{1/3}\). The optimality of the exponent 1/3 is not known. For the entire collection see [Zbl 1030.00027].

Recommendations

Cited in

(11)

This page was built for publication: The inverse problem for representation functions of additive bases

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q4462840)