Equivariant Fixed Point Index and Fixed Point Transfer in Nonzero Dimensions

Equivariant (K)-theory (19L47) Fixed-point and coincidence theorems (topological aspects) (54H25) Fixed points and coincidences in algebraic topology (55M20) Degree, winding number (55M25) Topological (K)-theory (55N15) Equivariant homology and cohomology in algebraic topology (55N91) Stable homotopy theory, spectra (55P42) Equivariant homotopy theory in algebraic topology (55P91) Equivariant homotopy groups (55Q91)

Review text

This paper extends \textit{A. Dold}'s work [Invent. Math. 25, 281-297 (1974; Zbl 0284.55007)] on the fixed point index and transfer to the equivariant setting. Given an \(\text{ENR}_{\mathcal B}\), \(\pi: E\to B\) and a map \(f: V\to E\), where \(V\) is open, \(f\) covers \(\pi\) and is compactly fixed \((\hbox{Fix}(f)=\{v\in V\mid f(v)=v\}\to B\) is proper), Dold defined a fixed point index for a generalized cohomology theory. In subsequent work [Math. Z. 148, 215-244 (1976; Zbl 0329.55007)] he defined a fixed point transfer from a ``cechification of the cohomology theory of the fixed point set to the cohomology of \(B\). The authors generalize these results for equivariant maps, that is, let \(G\) be a compact Lie group and let \(E\to B\) be a \(G\)-GNR\(_{\mathcal B}\). Let \(M\) and \(N\) be real \(G\) modules and let \(V\subset N\times E\) be open and \(G\) invariant and a map \(f: V\to M\times E\) which covers the projections \(N\times E\to B\) and \(M\times E\to B\) where \(f\) is equivariant and compactly fixed. They extend the fixed point transfer to an equivariant transfer, define an invariant fixed point index and prove the following Lefshetz-Hopf theorem (where \(\tau\), below, is the transfer): Theorem 1. Let \(f: N\times V'\to M\times E\) and let \(V'\subset E\) open and invariant, be a compactly fixed \(G\)-map over \(B\) such that its image \(f(N\times V')\) is contained in \(M\times K\) for some invariant subset \(K\subset V'\) which is proper and \(h_ G\)-flat over \(B\). If \(f\) is such that \(f_{-1}(0\times K)\cap)\cap(N\times K)\subset\mathbb{B}\times K\) for a ball \(\mathbb{B}\subset N\), then the composite homomorphism \[ \check h_ G(K)\cong \check h_ G(\mathbb{B}\times K)\buildrel\check j\over\longrightarrow \check h_ G(Fix(f))\buildrel\check\tau_ f\over\longrightarrow h_ G(B) \] where \(j:Fix(f)\subset\mathbb{B}\times K\), is given by the following trace formula \[ \check\tau\check j(a)=trace[x\mapsto a\smile(f\mid K)^ \vee(x)]\in h_ G^{| a|+[M]-[N]}(B), \] where we denote by \((f\mid K)^ \vee\) the following homomorphism of degree \([M]=[N]\in RO(G)\) defined by \(f: N\times K\to M\times K\) \[ \check h_ G(K)\cong\check h_ G((M,M- 0)\times K)\to\check h_ G((N,N-\mathbb{B})\times K)\cong \check h_ G \] (since \(f^{-1}(0\times K)\cap(N\times K)\subset\mathbb{B}\times K\), \(f\) maps \((N-\mathbb{B})\times K\) into \((M-0)\times K\). In other words, the image of \(a\in\check h_ G(K)\) under \(\check\tau_ f\check j\) is the trace of the composite \(h_ G(B)\)-endomorphism \[ \check h_ G(K)\buildrel (f\mid K)^ \vee\over\longrightarrow \check h_ G(K)\buildrel a\smile\over \longrightarrow \check h_ G(K). \] In particular, the index \(I(f)\) equals the trace of \((f\mid K)^ \vee: \check h_ G(K)\to\check h_ G(K)\).

Recommendations

Cites work

Cited in

(16)

This page was built for publication: Equivariant Fixed Point Index and Fixed Point Transfer in Nonzero Dimensions

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q3984300)