Hecke algebras and semigroup crossed product \(C^*\)-algebras (Q1306261): Difference between revisions

A subgroup \(\Gamma_0\) of a discrete group \(\Gamma\) is almost normal [after \textit{J.-B. Bost} and \textit{A. Connes}, Sel. Math., New Ser. 1, 412-457 (1995; Zbl 0842.46040)] if every \(\Gamma_0\)-orbit of \(\Gamma/\Gamma_0\) viewed as a left \(\Gamma_0\)-space is finite. The Hecke algebra \(H(\Gamma,\Gamma_0)\) of such a pair is the convolution algebra formed by all complex-valued, finitely-supported functions on the double coset space \(\Gamma_0\backslash \Gamma/\Gamma_0\). Every involutive homomorphism from \(H(\Gamma,\Gamma_0)\) to a \(C^\ast\)-algebra (for instance, the regular representation in \(\ell_2(\Gamma/\Gamma_0)\) [loc. cit.]) defines a norm on the Hecke algebra, and it is natural to ask under which conditions \(H(\Gamma,\Gamma_0)\) admits the maximal \(C^\ast\)-norm, because in this case one can form the corresponding completion and talk of the Hecke \(C^\ast\)-algebra of an almost normal subgroup pair. (Bost and Connes in their paper only considered the naturally defined, reduced Hecke \(C^\ast\)-algebra of such a pair.) Furthermore, assuming a pair \((\Gamma,\Gamma_0)\) admits a maximal \(C^\ast\)-algebra norm, can one describe its structure in familiar terms? The paper under review reports progress in both directions. In particular, the following is shown. Assume that \(\Gamma\) is a semidirect product group of the form \(G\ltimes N\), where \(\Gamma_0\trianglelefteq N\trianglelefteq\Gamma\), satisfying an additional restriction of a certain preorder completeness type. Then the (universal) Hecke \(C^\ast\)-algebra of the pair \((\Gamma,\Gamma_0)\) exists and is isomorphic to the semigroup crossed product \(C^\ast\)-algebra \(C^\ast(N/\Gamma_0)\rtimes T\), where \(T\) is a certain naturally defined subsemigroup of \(G\). In conclusion of the article, Hecke \(C^\ast\)-algebras are computed for some concrete examples of almost normal subgroup pairs.