Noether's problem for metacyclic \(p\)-groups (Q2497331)

Let \(K\) be a field and let \(G\) be a finite group. Let \(G\) act as a group of \(K\)-automorphisms of the rational function field \(L = K(x_g: g \in G)\) by \(g.x_h = x_{gh}\), and let the fixed subfield \(L^G\) of \(L\) be denoted by \(K(G)\). Noether's problem asks whether \(K(G)\) is rational (= purely transcendental) over \(K\). This problem arose in connection with attempts to realize a given \(G\) as the Galois group of a certain field extension (of \(\mathbb{Q}\)). A result of \textit{E. Fischer} in [Nachr. Kgl. Ges. Wiss. Göttingen, Math.-Phys. Kl. 1915, 77--80 (1915; JFM 45.0200.01)] establishes an affirmative answer in the case when \(G\) is abelian of exponent \(e\) and \(G\) contains a primitive \(e\)-th root of 1. A theorem of \textit{M. Hajja} in [J. Algebra 80, 295--305 (1983; Zbl 0544.20007)] extends this to the case when \(G\) is an extension of an abelian group by a cyclic group of order \(< 23\) and when \(K = \mathbb{C}\) (or a field that contains enough roots of 1). The first negative answer to Noether's problem was given by \textit{R. G. Swan} in his seminal paper in [Invent. Math. 7, 148--168 (1969; Zbl 0186.07601)], where he proved that \(\mathbb{Q} (G)\) is not rational if \(G\) is cyclic of order 47. Later works by V.~E.~Voskresenskii, S.~Endo and T.~Miyata, and H.~W.~Lenstra jun. completely settled the problem when \(G\) is abelian. The illusion, if any, that the non-rationality of \(K(G)\) can be solely due to the lack, in \(K\), of enough roots of 1 was dispelled by \textit{D. Saltman} in [Invent. Math. 77, 71--84 (1984; Zbl 0546.14014], where he proved that for any prime \(p\) and any field \(K\) with \(\text{char}\, K\neq p\) there is a meta-abelian group \(G\) of order \(p^9\) for which \(K(G)\) is non-rational. In the paper under review, the author considers the case when \(G\) is a meta-cyclic \(p\)-group of exponent \(p^e\), and establishes the rationality of \(K(G)\) in each of the cases (i) \(\text{char}\,K = p\) and (ii) \(\text{char}\,K\neq p\) and \(K\) contains a primitive \(p^e\)-th root of 1. He does this by reducing the problem to the case when \(G\) is a split meta-cyclic \(p\)-group and then by establishing a certain rationality criterion of the multiplicative invariants under a meta-cyclic group action, which is reminiscent of similar criteria established earlier by H.~W.~Lenstra jun. and by M. Hajja. Besides being a valuable addition to the literature on Noether's problem, the paper under review contains a well-written introduction that summarizes the most important results known so far.