English

Sums along the edges of bounded degree graphs

Combinatorics 2025-08-04 v2

Abstract

Let GG be a graph on nn vertices and (H,+)(H,+) be an abelian group. What is the minimum size SH(G){\sf S}_H(G) of the set of all sums A(u)+A(v)A(u)+A(v) over all injections A:V(G)HA:V(G)\to H? In 2012, the first author, Angel, the second author, and Lubetzky proved that, for expander graphs and H=ZH=\mathbb{Z}, this minimum is at least Ω(logn)\Omega(\log n), and this bound is tight -- there exists a regular expander GG with SZ(G)=O(logn){\sf S}_{\mathbb{Z}}(G)=O(\log n). We prove that, for every constant d3d\geq 3, the random dd-regular graph Gn,d\mathcal{G}_{n,d} has significantly larger sum-sets: with high probability, for every abelian group HH, SH(Gn,d)=Ω(n12/d){\sf S}_H(\mathcal{G}_{n,d})=\Omega(n^{1-2/d}). In particular, this proves that, for every ε>0\varepsilon>0, there exists a regular graph with O(n)O(n) edges and with sum-sets of size at least n1εn^{1-\varepsilon}, for all abelian groups. The bound SH(Gn,d)=Ω(n12/d){\sf S}_H(\mathcal{G}_{n,d})=\Omega(n^{1-2/d}) is tight up to a polylogarithmic factor: We show that, for every 3dlnn/lnlnn3\leq d\leq \ln n/ \ln \ln n, there exists an abelian group HH such that, for every graph GG on nn vertices with maximum degree at most dd, SH(G)n12/d(logn)O(1){\sf S}_H(G) \leq n^{1-2/d}(\log n)^{O(1)}. We also prove that, for dln2nd\gg\ln^2 n, with high probability, for every abelian group HH, SH(Gn,d)=n(1o(1)){\sf S}_H(\mathcal{G}_{n,d})=n(1-o(1)) and determine the second-order term, up to a polylogarithmic factor.

Keywords

Cite

@article{arxiv.2507.01138,
  title  = {Sums along the edges of bounded degree graphs},
  author = {Noga Alon and Itai Benjamini and Georgii Zakharov and Maksim Zhukovskii},
  journal= {arXiv preprint arXiv:2507.01138},
  year   = {2025}
}

Comments

This is a substantial extension of the first version with a different set of authors and title. In particular, the new version contains an upper bound on sum-sets of random regular graphs which is tight, up to a polylogarithmic factor