Unramified extensions over low degree number fields (Q2307444)

Results concerning the inverse Galois problem for extensions of quadratic global fields are grouped around two questions. Question 1. Does every finite group appear as the Galois group of an unramified extension of some quadratic number field? The formulation of the second question requires two definitions. Let \(G\) be a finite group. Then \[ge(G) := \min_{S} \mathrm{lcm} \{\mathrm{ord} (x)\mid x \in S\},\] where the minimum of the least common multiples is taken over all generating subsets \(S\) of the group \(G\). Furthermore, we define \(e(\mathbb{Q}, G)\) as the minimal degree of a number field \(K\) which admits an unramified \(G\)-extension defined over \(\mathbb{Q}\) by a tame extension. Question 2. Is it true that \(e(\mathbb{Q}, G) = ge(G)\)? In the function field case, the authors consider extensions of \(k(x)\) and prove that the answer to both questions is affirmative if the coefficient field \(k\) is ample. The main tools used in the proofs are Abhyankar's lemma on ramification in the compositum of two extension fields and a theorem by [\textit{S. Beckmann}, J. Reine Angew. Math. 419, 27--53 (1991; Zbl 0721.11052)] on the specializations of a branched covering of the projective line to points, where both the covering and the points are defined over a number field. The results support an affirmative answer to question 2 and provide a criterion for the equality. In particular, proposition 4.5 proves the equality in question 2 for a list of eight nonabelian almost simple groups of small order. These eight groups, which include the smallest sporadic simple group \(M_{11}\), serve as examples for the criterion the authors prove. Further applications are possible for cases where certain group-theoretic conditions are satisfied.