On the Erdős-Volkmann and Katz-Tao ring conjectures (Q2504001)

The author proves the so-called 'discretized ring conjecture' of \textit{N. H. Katz} and \textit{T. Tao} [New York J. Math. 7, 149--187 (2001; Zbl 0991.28006)]. We say that a bounded set \(A \subseteq \mathbb R\) is a \((\delta, \sigma)_1\)-set if \(A\) is a union of intervals of length comparable to \(\delta\) and if for any interval \(I \subseteq \mathbb R\) of length \(r \in [\delta, 1]\), \[ | A \cap I| < \Big({r \over \delta}\Big)^{1-\sigma} \delta^{1-\varepsilon}, \] for some small \(\varepsilon > 0\). It is shown that if \(A\) is a \((\delta, \sigma)_1\)-set with \(0 < \sigma < 1\) and \(| A| > \delta^{\sigma+\varepsilon}\), then \[ | A+A| + | A.A| > \delta^{\sigma - c}, \] where \(c>0\) depends only on \(\sigma\). This was conjectured by \textit{Katz} and \textit{Tao} [loc.cit.] in the case \(\sigma = 1/2\). As a consequence of the above result, the author answers a question of \textit{P. Erdős} and \textit{B. Volkmann} [J. Reine Angew. Math. 221, 203--208 (1966; Zbl 0135.10202)] by proving that a Borel subring of \(\mathbb R\) cannot have Hausdorff dimension strictly between \(0\) and \(1\). The result is also applied to the Falconer distance problem, where it implies that if \(K \subseteq \mathbb R^2\) is a compact set of Hausdorff dimension at least \(1\), then the set of distances \(| K-K| \) has Hausdorff dimension at least \(1/2 + c_0\), where \(c_0>0\) is a universal constant. The Falconer conjecture claims that \(c_0 = 1/2\). Another application of the main result is to the Furstenberg conjecture. Let \(K \subseteq \mathbb R\) be a compact set such that for every direction \(\omega \in S^1\), there is a line segment with direction \(\omega\), intersecting \(K\) in a set of Hausdorff dimension at least \(1/2\). It follows from the main result of the present paper, that the Hausdorff dimension of \(K\) is at least \(1 + c\), where \(c>0\) is a universal constant. The connections between the discretized ring conjecture, the Falconer distance problem and the Furstenberg conjecture were explored in the paper by \textit{N. H. Katz} and \textit{T. Tao} [loc.cit.]. A different proof of the Erdős--Volkmann conjecture avoiding the discretized ring conjecture was given by \textit{G. A. Edgar} and \textit{C. Miller} [Proc. Am. Math. Soc. 131, No. 4, 1121--1129 (2003; Zbl 1043.28003)]. The main result of the present paper is much more general. The proof is very technical, and rather difficult to read. Methods from harmonic analysis are combined with a quantitative version of the Balog-Szemerédi theorem due to \textit{W. T. Gowers} [Geom. Funct. Anal. 8, No. 3, 529--551 (1998; Zbl 0907.11005)] as well as various measure theoretical and combinatorial arguments.