A conjecture about character degrees of solvable groups (Q1176648): Difference between revisions

For a finite solvable group \(G\) let \(\rho(G)\) be the set of primes which divide some irreducible (classical) character degree of \(G\) and let \(\sigma(G)\) be the maximum number of different primes which divide any one irreducible character degree. Now, Huppert asked the question: Is \(|\rho(G)|\leq 2\sigma(G)\)? For \(\sigma(G)=1\), the affirmative answer follows from the thesis of \textit{O. Manz} [J. Algebra 94, 211-255 (1985; Zbl 0596.20007)]. The best general result so far known is \(|\rho(G)|\leq 3\sigma(G)+32\) [\textit{D. Gluck} and \textit{O. Manz}, Bull. Lond. Math. Soc. 19, 431-437 (1987; Zbl 0627.20007)]. In the paper under review the author gives an affirmative answer to Huppert's question in two special situations, namely when \(\sigma(G)=2\) and secondly, when all irreducible character degrees of \(G\) are squarefree. In the first case the proof splits up into three subcases according to the number of primes \(p\) for which \(O_ p(G)\) is nonabelian. (Note that there are at most two such primes.) Each case then reduces to a nilpotent by metabelian group for which the character degree analysis can be done. In the later case, the groups that the author has to deal with are known by \textit{B. Huppert} and \textit{O. Manz} [Arch. Math. 45, 125-132 (1985; Zbl 0574.20006)]. They are nilpotent by metacyclic.