Noether's problem for four- and five-dimensional linear actions (Q994273)

Let \(G\) be a permutation group contained in the symmetric group \(S_n\) for some integer \(n\), and \(K\) be a field. Then \(G\) acts naturally on the rational function field \(K(x_1, x_2, \cdots, x_n)\) by \(K\)-automorphisms. Noether's problem asks whether the fixed subfield \(K(x_1, x_2, \cdots, x_n)^G=\{ f \in K(x_1, x_2, \cdots, x_n): \sigma (f) = f \text{ for all } \sigma \in G \}\) is rational (i.e. purely transcendental) over \(K\). More generally, if \(G\) is a finite group, \(K\) is a field and \(G \rightarrow \mathrm{GL}(V)\) is a faithful representation of \(G\) where \(V\) is a finite dimensional vector space over \(K\), Noether's problem asks whether the fixed subfield \(K(V)^G\) is rational over \(K\). It is known that, if \(G \rightarrow \mathrm{GL}(V_i)\) are two faithful representations of \(G\) for \(i=1,2\), then \(K(V_1)^G\) is stably rational over \(K\) if and only if so is \(K(V_2)^G\) over \(K\). Because there are so many faithful representations of a finite group \(G\), a convenient practice for many people (e.g. Lenstra, Saltman, etc.) is to consider the faithful representation \(G \rightarrow \mathrm{GL}(V)\) when \(V= V_{\mathrm{reg}}\) is the regular representation of the group \(G\). Thus people study whether \(K(V_{\mathrm{reg}})^G\) is rational over \(K\) and call it the Noether's problem. The purpose of this paper is to study Noether's problem when the group \(G\) is any finite \(2\)-group contained in \(\mathrm{GL}(V)\) where \(V\) is a vector space over \(Q\) with dimension equal to \(4\) or \(5\), and the faithful representation of the group is just the natural embedding of \(G\) into \(\mathrm{GL}(V)\). The author proves that \(Q(V)^G\) is rational for any such group \(G\), except when \(G \simeq C_8\), or \(C_8 \times C_2\). When \(G \simeq C_8\), or \(C_8 \times C_2\), \(Q(V)^G\) is not rational over \(Q\) by a result of Voskresenskii (and Lenstra also). Here is the main idea of the proof. From the book ``Crystallographic groups of four-dimensional space'' by H. Brown etc. one can find a list of all these subgroups up to conjugation. There are \(67\) such groups in \(GL_4(Q)\). As to the subgroups in \(GL_5(Q)\), since they are reducible, the rationality problem of \(Q(V)^G\) when \(\dim V =5\) can be reduced to that of \(Q(V)^G\) when \(\dim V =4\) by the standard technique. Hence it remains to consider the rationality problem for these \(67\) groups. It is found by the author that \(63\) groups among the \(67\) groups can be solved at one stroke (see Theorem 14). The answers to the remaining \(4\) group were already known by other authors.