Wiki
Wiki

Research notes on every problem, and a library of the papers behind them. Built from the open erdos repository.

Updated


Claim. Let A⊂{1,…,N}A\subset\{1,\ldots,N\} be a Sidon set with ∣A∣=k≥N1/2−l(N)\lvert A\rvert=k\ge N^{1/2}-l(N), where l=o(N1/2)l=o(N^{1/2}), and let m=o(N1/2)m=o(N^{1/2}). Writing a(x)a(x) for the number of elements of AA congruent to xx modulo mm, Theorem 2 of [[../library/additive_bases/kolountzakis_1999_uniform_distribution_residue_classes_dense_sets/_index|the paper]] bounds the ℓ2\ell^2 discrepancy ∥a(x)−k/m∥2\lVert a(x)-k/m\rVert_2 by CN3/8/m1/4CN^{3/8}/m^{1/4} when l≤N1/4m1/2l\le N^{1/4}m^{1/2} and by CN1/4l1/2/m1/2CN^{1/4}l^{1/2}/m^{1/2} otherwise, and its remarks deduce a(x)=k/m+o(k/m)a(x)=k/m+o(k/m) uniformly in xx for m=o(N1/6)m=o(N^{1/6}) in the first range and m=o(N1/2/l)m=o(N^{1/2}/l) in the second. For a fixed modulus and ∣A∣∼N1/2\lvert A\rvert\sim N^{1/2} this recovers Lindström's equidistribution of AA, and for constant mm and l≤CN1/4l\le CN^{1/4} it bounds the ℓ2\ell^2 discrepancy by CmN3/8C_mN^{3/8}, the error Lindström obtained only for m=2m=2 and ∣A∣≥N1/2\lvert A\rvert\ge N^{1/2}; the statement for A+AA+A follows by the Sidon property as the page of [[problems/additive_bases/E0154/claims/1998_04_01_lindstrom|Lindström's claim]] explains. The method is analytic: the input is the author's earlier estimate for nonnegative cosine polynomials with distinct integer frequencies.

Depends on. [[problems/additive_bases/E0154/claims/1998_04_01_lindstrom|Lindström's page]] for the deduction of the sumset statement from the equidistribution of AA; the equidistribution itself is the paper's.

Acceptance. Refereed: M. N. Kolountzakis, On the uniform distribution in residue classes of dense sets of integers with distinct sums, J. Number Theory 76 (1999), no. 1, 147–153; the page name uses the date of the first version of the preprint, arXiv:math/9808061, 1998-08-14. Reviewed: the site's curator, T. F. Bloom, records the problem as proved at erdosproblems.com on Lindström's theorem and this strengthening, which is the site's acceptance. No Lean formalization of this quantitative statement is recorded; the formalizations linked from Lindström's page cover the qualitative statement.