The coarse geometric Novikov conjecture for subspaces of non-positively curved manifolds (Q2372180): Difference between revisions

A discrete metric space \(\Gamma\) is said to have bounded geometry if for any \(r>0\) there is a natural number \(N\) such that any ball of radius \(r\) in \(\Gamma\) contains at most \(N\) elements. For each integer \(d\geq0\), the Rips complex \(P_d(\Gamma)\) is defined to be the simplicial polyhedron in which the set of vertices is \(\Gamma\), and a finite subset \(\{\gamma_1, \gamma_2, \dots, \gamma_n\}\subset\Gamma\) spans a simplex if and only if \(d(\gamma_i, \gamma_j)\leq d\) for all \(0\leq i, j\leq n\). The authors' main theorem: If \(\Gamma\) admits a coarse embedding into a simply connected complete Riemannian manifold \(M\) of non-positive sectional curvature, then the index map \(\text{ind}: \lim_{d\to\infty} K_*(P_d(\Gamma))\to \lim_{d\to\infty} K_*(P_d(\Gamma))\cong K_*(C^*(\Gamma))\) is injective. Let \(\mathcal A=C_0(M, \text{Cliff}(TM))\) denote the \(C^*\)-algebra of continuous functions on \(M\) which have value \(a(x)\in \text{Cliff}(T_xM)\) at each point \(x\in M\) and vanish at infinity, where \(\text{Cliff}(T_xM)\) is the complexified Clifford algebra of the tangent space \(T_xM\) at \(x\) with respect to the inner product on \(T_xM\) given by the Riemannian structure of \(M\). The authors introduce the notions of twisted Roe algebras \(C^*(P_d(\Gamma)), \mathcal A)\) and of twisted localization algebras \(C^*_L(P_d(\Gamma)), \mathcal A)\) for a bounded geometric space which admits a coarse embedding into \(M\) and at first show that the evaluation map is an isomorphism \(\lim_{d\to\infty}K_*(C^*_L(P_d(\Gamma)), \mathcal A) \!\overset\cong{\to} \lim_{d\to\infty} K_*(C^*(P_d(\Gamma)), \mathcal A)\) (Theorem 3,3) by parallel arguments of \textit{G. Yu} [Invent. Math. 139, No. 1, 201--240 (2000; Zbl 0956.19004)]. The authors construct uniformly almost flat Bott generators of \textit{G. Kasparov} and \textit{G. Yu} [Adv. Math. 206, 1--56 (2006; Zbl 1102.19003)] for \(M\), define a Bott map \(K_*(C^*(P_d(\Gamma))\to K_*(C^*(P_d(P_d(\Gamma)), \mathcal A)\) and another Bott map \(K_*(C^*_L(P_d(\Gamma))\to K_*(C^*_L(P_d(\Gamma)), \mathcal A)\), then show the latter is an isomorphism (Theorem 5.4). Since \(\text{ind}_L: \lim_{d\to\infty} K_*(P_d(\Gamma))\to \lim_{d\to\infty} K_*(C_*(P_d(\Gamma)))\) is an isomorphism by \textit{G. Yu} [K-Theory 11, No. 4, 307--318 (1997; Zbl 0888.46047)], the main theorem is proved.