Let $G$ be an arbitrary (discrete) group. It acts by left translation on $\ell^\infty(G)$. The uniform Roe algebra of $G$ is defined as the crossed product $\ell^\infty (G) \rtimes_{\mathrm{red}}G$.

Elmar Thoma has shown (Thoma, E., Eine Charakterisierung diskreter Gruppen vom Typ I, Invent. Math. 6, 190-196 (1968). ZBL0169.03802.) that the reduced group $C^*$-algebra $C_{\mathrm{red}}^*(G)$ of a group $G$ is a type I $C^*$-algebra (i.e. there exists some $n\in \mathbb{N}$ such that for every irreducible representation $\pi$ of $C_{\mathrm{red}}^*(G)$ the dimension of the corresponding Hilbert space is less or equal to $n$) if and only if $G$ is virtually abelian.

I'm wondering if for virtually abelian groups the corresponding uniform Roe algebra must also be of type I. Does a result into that direction exist?


It is not Type I in general. Probably it is not Type I whenever $G$ is infinite. Here is an argument when $G=\mathbb{Z}.$

Consider the projections in $\ell^\infty(\mathbb{Z})$ defined by characteristic functions for the following sets $\{ 2^n\mathbb{Z}+k: n\geq 1,0\leq k <2^n \}.$ Let $A$ be the C*-algebra generated by these projections. Then $A=C(X)$ where $X=\{ 0,1 \}^\mathbb{N}$. Moreover the $\mathbb{Z}$-action leaves $A$ invariant and defines the odometer action on $\{ 0,1 \}^\mathbb{N}.$ The odometer action is minimal and therefore $A\rtimes \mathbb{Z}\subseteq \ell^\infty(\mathbb{Z})\rtimes \mathbb{Z}$ is a simple, unital, infinite dimensional C*-algebra and therefore not Type I (hence it can't be contained in any Type I C*-algebra).

Side Note: Zeller-Meier (1968) gave a dynamical characterization of those $\mathbb{Z}$-actions on compact spaces whose crossed product is Type 1--$C(X)\rtimes \mathbb{Z}$ is Type I precisely when every orbit is discrete in the relative topology. Having a minimal subsystem (as above) is then about as far away as you can get from Zeller-Meier's condition

| cite | improve this answer | |
  • $\begingroup$ Every infinite finitely generated group G contains $\mathbb{Z}$ isometrically. Hence, uniform Roe algebra of G contains uniform Roe algebra of $\mathbb{Z}$ as a C*-subalgebra. Moreover, type I passes to C*-subalgebras. $\endgroup$ – m07kl Oct 13 '19 at 12:23

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.