A decomposition formula for equivariant stable homotopy classes (Q1039204)

Let \(G\) be a compact Lie group and \(X, Y\) two metric \(G\)-spaces. For \(k \in \mathbb{Z}\) denote by \(\{X, Y\}^k_G\) the group of stable homotopy classes represented by \(G\)-maps of pairs \[ \alpha : (\mathbb{R}^n\times V, \mathbb{R}^n\times V-0)\times X \to (\mathbb{R}^{n+k}\times V, \mathbb{R}^{n+k}\times V-0)\times Y \] for some orthogonal representation \(V\) of \(G\). In this paper the authors assume that \(X\) is a trivial \(G\)-space and give a decomposition of \(\{X, Y\}^k_G\) as a direct sum of subgroups \( \{X, Y\}^k_{(H)}\) of fixed orbit types (Theorem 2.6). The definition of these subgroups depends on a formula due to the second author [Topology Appl. 58, No. 2, 181--191 (1994; Zbl 0830.55011)], which gives a unique factorization of elements of \(\{X, Y\}^k_G\) as a composition. This formula is derived from a particular kind of \(G\)-map, called an equivariant fixed-point situation over \(X\), by considering its transfer. More precisely, given an equivariant map \(f\) from an open \(G\)-subset \({\mathcal U} \subset \mathbb{R}^n\times V \times X\) to \(\mathbb{R}^{n+k}\times V \times X\) commuting with the two canonical projections onto \(X\), its transfer can be interpreted as a stable map \[ \tau(f) : (\mathbb{R}^n\times V, \mathbb{R}^n\times V-0)\times X \to (\mathbb{R}^{n+k}\times V, \mathbb{R}^{n+k}\times V-0)\times {\mathcal U}. \] Then the formula above asserts that any class \(\{\alpha \} \in \{X, Y\}^k_G\) (assuming \(\alpha\) is represented as above) can be factored as the composition in the form \(\{\alpha \}=(1\times\varphi)\circ\tau(f)\) where \(\varphi : {\mathcal U} \to Y\) is a nonstable equivariant map and 1 is the identity on \(\mathbb{R}^{n+k}\times V\). This expression allows one to define \(\{X, Y\}^k_{(H)}\) to be the subgroup of those classes \(\{\alpha \}=(1\times\varphi)\circ\tau(f)\) where \(f\) is a conormal map such that \(\text{Fix}(f) \subset {\mathcal U}_{(H)}\), \({\mathcal U}_{(H)}\) being the set consisting of the points of \({\mathcal U}\) with isotropy group conjugate to \(H\). The proof of the theorem presented above is done by applying arguments similar to those used in the previous paper by the authors [J. Lond. Math. Soc., II. Ser. 69, No.~1, 214--230 (2004; Zbl 1061.55002)]. In fact the approach taken here is to represent \(\tau(f)\) as a sum \(\tau(f)=\Sigma_{(H)}\tau(f_{(H)})\) where \(f_{(H)}\) denotes the restriction of \(f\) to a certain neighbourhood of \(\text{Fix}(f|_{{\mathcal U}_{(H)}})\). This enables one to obtain a direct sum decomposition \(\{\alpha \}=\Sigma_{(H)}\tau(\alpha_{(H)})\) where \(\{\alpha_{(H)} \} \in \{X, Y\}^k_{(H)}\), and thereby to complete the proof.