The Gromov-Lawson-Rosenberg conjecture for groups with periodic cohomology (Q1383637)

The conjecture of the title concerns the question of whether a given manifold \(M\) admits a metric of positive scalar curvature (PSC). \textit{A. Lichnerowicz} [C. R. Acad. Sci., Paris 257, 7-9 (1963; Zbl 0136.18401)] showed that there are some manifolds without a PSC metric. Rosenberg constructed an invariant \(\alpha(M)\), taking values in the real \(K\)-theory of an object built from the fundamental group of \(M\), which vanishes on manifolds which do admit a PSC metric. Modifying a conjecture of Gromov-Lawson, he conjectured that this invariant vanishes if and only if \(M\) admits a PSC metric, subject to certain conditions on \(M\). Namely, \(M\) is assumed to be closed, connected, smooth, spin and of dimension at least \(5\). This conjecture has been verified for such manifolds whose fundamental group is either trivial, odd order cyclic or of order 2, or one of a small number of infinite groups. The main theorem of this paper verifies the conjecture for manifolds whose fundamental groups are finite and have periodic cohomology. To prove this, it is sufficient to check the conjecture for cyclic groups of order \(2^k\) (\(k \geq 2\)) and generalized quaternion groups. The theorem then follows from known results since the finite groups with periodic cohomology are precisely those whose \(p\)-Sylow subgroups are cyclic for odd \(p\) and cyclic or (generalized) quaternion groups for \(p=2\). To give an idea of the techniques used, we list some of the section headings: \(ko\)-homology calculations, a stable splitting of \(BQ_l\), analytic results concerning the eta invariant, the range of the eta invariant for cyclic and generalized quaternion groups.