Infinite unrestricted sumsets in subsets of abelian groups with large density
Dimitrios Charamaras, Ioannis Kousek, Andreas Mountakis, Tristán Radić · 2025
Read Infinite unrestricted sumsets in subsets of abelian groups with large density by Dimitrios Charamaras, Ioannis Kousek, Andreas Mountakis, Tristán Radić (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2504-08649.txt.
Introduction
Charamaras and co-authors find vast sum sets in dense parts of swap groups. A swap group here is a count-able set where sums swap free. Two group traits frame the work: a thin twin set and a thin twin map. The team sets low bounds on how dense a part must be. Past that bound, the part holds a sum set of a set form. The form sums two picks from one vast set B in the group.
A shift t may lead the sum set in one of the two forms. Both group traits prove a must for the claims to hold. The team also marks which spread chains let the claim run. Their low bounds prove best in a strong sense in the work. No less dense part can dodge the sum set by their proof.
Catalog record
| CREATOR | Dimitrios Charamaras, Ioannis Kousek, Andreas Mountakis, Tristán Radić |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 120 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2504-08649
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Infinite unrestricted sumsets in subsets of abelian groups with large density Dimitrios Charamaras, Ioannis Kousek, Andreas Mountakis, Tristán Radić Abstract . Let ( G , + ) (G,+) be a countable abelian group such that the subgroup { g + g : g ∈ G } \{g+g\colon g\in G\} has finite index and the doubling map g ↦ g + g g\mapsto g+g has finite kernel. We establish lower bounds on the upper density of a set A ⊂ G A\subset G with respect to an appropriate Følner sequence, so that A A contains a sumset of the form { t + b 1 + b 2 : b 1 , b 2 ∈ B } \{t+b_{1}+b_{2}\colon b_{1},b_{2}\in B\} or { b 1 + b 2 : b 1 , b 2 ∈ B } \{b_{1}+b_{2}\colon b_{1},b_{2}\in B\} , for some infinite B ⊂ G B\subset G and some t ∈ G t\in G . Both assumptions on G G are necessary for our results to be true. We also characterize the Følner sequences for which this is possible. Finally, we show that our lower bounds are optimal in a strong sense. 1. Introduction In [ 12 ] , Kra, Moreira, Richter and Robertson resolved a longstanding conjecture of Erdős (see for example [ 6 , Page 305] ) via the following theorem. Theorem A . [ 12 , Theorem 1.2] For any A\subset{\mathbb{N}} with positive upper Banach density there exist an infinite set B\subset A and some t\in{\mathbb{N}} such that B\oplus B:=\{b_{1}+b_{2}\colon b_{1},b_{2}\in B,\ b_{1}\neq b_{2}\}\subset A-t. It is natural to ask whether Theorem A could be extended to other countable amenable (semi)groups. This is explored in [ 4 ] , where the first and third authors establish an extension of Theorem A for a wide class of such groups. This class includes all finitely generated nilpotent groups, and all abelian groups (G,+) with the property that 2G:=\{2g\colon g\in G\}, where 2g:=g+g , has finite index. Throughout, unless explicitly stated otherwise, all the groups we consider are countable. Given an abelian group G , a sequence \Phi=(\Phi_{N})_{N\in{\mathbb{N}}} of finite subsets of G is called a Følner sequence if for any g\in G , \lim_{N\to\infty}\frac{|\Phi_{N}\cap(g+\Phi_{N})|}{|\Phi_{N}|}=1. We denote the nonempty set of all Følner sequences in G by \mathcal{F}_{G} . The upper density of A\subset G with respect to \Phi\in\mathcal{F}_{G} is defined as \overline{\mathop{}\!\mathrm{d}}_{\Phi}(A):=\limsup_{N\to\infty}\frac{|A\cap\Phi_{N}|}{|\Phi_{N}|}. We say that A has positive upper Banach density if \overline{\mathop{}\!\mathrm{d}}_{\Phi}(A)>0 for some \Phi\in\mathcal{F}_{G} . Then [ 4 , Corollary 1.13] asserts that if (G,+) is a countable abelian group with index [G:2G]<\infty and A\subset G has positive upper Banach density, there exist an infinite set B\subset A and some t\in G such that B\oplus B\subset A-t . Another natural question arising from Theorem A is whether the restriction b_{1}\neq b_{2} in the sumset can be removed. It turns out that this restriction is necessary, as one can construct a set A\subset{\mathbb{N}} of full upper Banach density that contains no set of the form t+B+B=\{t+b_{1}+b_{2}:b_{1},b_{2}\in B\} where B\subset{\mathbb{N}} is infinite and t\in{\mathbb{N}} (see [ 14 , Example 2.3] ). It would thus be interesting to explore what – if any at all – density assumptions on the set A\subset{\mathbb{N}} would allow one to drop the restriction b_{1}\neq b_{2} . This was studied by the second and fourth authors in [ 11 ] . There, it is shown that for any A\subset{\mathbb{N}} such that \overline{\mathop{}\!\mathrm{d}}_{([1,N])}(A)>2/3 , there are an infinite set B\subset{\mathbb{N}} and some t\in\{0,1\} such that t+B+B\subset A . We remark that taking the density with respect to the initial Følner sequence N\mapsto[1,N] in {\mathbb{N}} is essential, as it is clear from [ 14 , Example 2.3] that not all Følner sequences can be used to guarantee such density threshold values. For additional results and open problems on infinite sumsets, we refer the reader to [ 2 , 9 , 10 , 13 , 14 , 15 , 17 ] . The main aim of the preceding discussion is to set the stage for the following – a posteriori natural – questions, which we address in this paper: Let G be an abelian group, such that [G:2G]<\infty . (a) Can we find a Følner sequence \Phi and a constant c=c(G,\Phi)>0 such that any set A\subset G with \overline{\mathop{}\!\mathrm{d}}_{\Phi}(A)>c contains an unrestricted sumset of the form t+B+B for some infinite set B\subset G and some t\in G ? (b) Can we answer (a) in an optimal way? We refer to this problem as the unrestricted B+B problem . Throughout, G denotes a countable abelian group and D denotes the doubling map D:G\to G,D(g)=2g . In addition, given a set A\subset G , we denote the set D(A) by 2A and the set D^{-1}(A) by A/2 . As we alluded to earlier, not all Følner sequences have density thresholds for the unrestricted B+B problem. However, we are able to pinpoint structural properties of a Følner sequence that allow for this to happen and we describe those in the next definition. Definition 1.1 . Let G be a countable abelian group. (i) We define the doubling ratio of a Følner sequence \Phi=(\Phi_{N})_{N\in{\mathbb{N}}} in G as (1.1) \alpha_{\Phi}=\liminf_{N\to\infty}\frac{|\Phi_{N}/2\cap\Phi_{N}|}{|\Phi_{N}|}. Whenever \alpha_{\Phi}>0 , we say that the Følner sequence \Phi is quasi-invariant with respect to doubling with ratio \alpha_{\Phi} , and we abbreviate this as q.i.d. with ratio \alpha_{\Phi} . (ii) We define the group doubling ratio as \alpha_{G}=\sup\{\alpha_{\Phi}\colon\Phi\in\mathcal{F}_{G}\} , where, as noted before, \mathcal{F}_{G} denotes the set of all Følner sequences in G . As we show in Section 6 , any abelian group G with \ell=[G:2G]<\infty and r=|\ker(D)|<\infty admits Følner sequences that are quasi-invariant with respect to doubling. More precisely, we prove that \alpha_{G}=\min\{1,\frac{r}{\ell}\} and that the value \alpha_{G} is attained, i.e., there is a Følner sequence \Phi in G so that \alpha_{\Phi}=\alpha_{G} . Next, we state our main theorem, which asserts that the Følner sequences defined in Definition 1.1 possess the necessary structural properties to provide an affirmative answer to question (a) . This allows us to resolve this question for all abelian groups G with [G:2G]<\infty and |\ker(D)|<\infty . Theorem 1.2 . Let (G,+) be a countable abelian group with \ell=[G:2G]<\infty and r=|\ker(D)|<\infty . Let A\subset G and \Phi be any Følner sequence in G that is quasi-invariant with respect to doubling with ratio \alpha_{\Phi} . Then the following hold: (1) If \displaystyle\overline{\mathop{}\!\mathrm{d}}_{\Phi}(A)>1-\frac{\ell\alpha_{\Phi}}{\ell+r}, then there exists an infinite set B\subset G and some t\in G such that t+B+B\subset A . (2) If \displaystyle{\overline{\mathop{}\!\mathrm{d}}_{\Phi}(A)>1-\frac{\alpha_{\Phi}}{\ell+r}}, then there exists an infinite set B\subset G such that B+B\subset A . (3) If \displaystyle\overline{\mathop{}\!\mathrm{d}}_{\Phi}(A\cap 2G)>\frac{1}{\ell}-\frac{\alpha_{\Phi}}{\ell+r}, then there exists an infinite set B\subset G such that B+B\subset A . The statements (1), (2) and (3) of Theorem 1.2 are logically equivalent; this is proved in Section 2.1 . Hence, it suffices to establish only one of the statements. In Section 2.3 , we reformulate statement (2) into a dynamical statement, which we then prove in Section 3 , concluding the proof of Theorem 1.2 . A useful interpretation of the equivalence between ( 2 ) and ( 3 ) in Theorem 1.2 is that in order to find patterns of the form B+B in A it suffices to check how many even elements the set