On the ubiquity of Sidon sets

From MaRDI portal





A subset \({\mathcal A}\) of \(\mathbb{N}\) is called a \(B_h[g]\) set if, for any positive integer \(n\), there are at most \(g\) solutions to the equation \(n=a_1+\cdots+a_h\) with \(a_1,\dots,a_h\in{\mathcal A}\) and \(a_1\leq\cdots \leq a_h\). The most famous example is the one of Sidon sets (corresponding to the parameters \(g=1\), \(h=2)\). The main result of the paper under review is that, when \(n\) tends to infinity, almost all subsets of \(\{1,2,\cdots,n\}\) with a cardinality \(o(n^{g/2(g+1)})\) are \(B_2[g]\) sets. Notice that a \(B_2[g]\) set included in \(\{1,2,\dots, n\}\) cannot have more than \(c_g\sqrt n\) elements while \(g/2(g+1)\) tends to 1/2 as \(g\) tends to infinity. A result (Theorem 6) of the same flavour concerning \(B_h[1]\) sets is also obtained. For the entire collection see [Zbl 1030.00027].











This page was built for publication: On the ubiquity of Sidon sets

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q4462841)