On the minimal ramification problem for semiabelian groups (Q716052)

For a finite group \(G\), let \(\mathrm{d}(G)\) be the smallest integer for which \(G\) is the normal closure of a subset \(S\) with \(\mathrm{d}(G)\) elements. One observes that if there exists a tame \(G\)-Galois extension \(K/\mathbb Q\), then at least \(\mathrm{d}(G)\) rational primes ramify in \(K\). It is an open question whether there exists such an extension with exactly \(\mathrm{d}(G)\) ramifying primes. This minimal ramification problem was solved by \textit{H. Kisilevsky} and \textit{J. Sonn} [Compos. Math. 146, No. 3, 599--606 (2010; Zbl 1197.11150)], using a result by \textit{D. Neftin} [``On semiabelian \(p\)-groups'', preprint, 2009, \url{arXiv:0908.1472v2}], for finite semiabelian \(p\)-groups, i.e. \(p\)-groups in the minimal family that contains \(\{1\}\) and is closed under quotients and semidirect products with finite abelian groups. In this paper the result is generalized to arbitrary finite semiabelian groups by means of a wreath product length \(\mathrm{wl}(G)\): when this coincides with \(\mathrm{d}(G)\) the minimal ramification problem is solved. In particular this includes all finite nilpotent semiabelian groups. In the last section the authors examine some arithmetic consequences that a general positive solution to the problem would have.