Here is a question from the proof of the "QD C*-algebra's representation theorem" in P245 of book "C*-algebras and Finite-Dimensional Approximations" by Nate and Taka.

For a separable unital C*-algebra $A$, let $\phi_{n}: A\rightarrow M_{k(n)}(\mathbb{C})$ be u.c.p maps which are asymptotically multiplicative (i.e., $||\phi_{n}(ab)-\phi_{n}(a)\phi_{n}(b)||\rightarrow 0$ for all $a, b\in A$) and isometric (i.e., $||a||=\lim_{n\rightarrow\infty}||\phi_{n}(a)||$ for all $a\in A$). Consider the u.c.p. map $$\Phi: A\rightarrow \prod_{n=1}^{\infty}M_{k(n)}(\mathbb{C})\subset B(\bigoplus_{n=1}^{\infty}\ell_{k(n)}^{2}),~~\Phi(a)=\bigoplus_{n=1}^{\infty}\phi_{n}(a).$$

Then, $\Phi$ is a faithful representation modulo the compact.

My question are:

  1. How to verify $\Phi$ is a representation modulo the compact? (I think the "faithful" is easy from the isometric.)

  2. In the proof, the author says $\Phi$ obviously has finite-rank projections which commute with its image and tend strongly to one. What are the "obvious" finite-rank projections?

Now, I attach the definition of representation modulo the compact in P19 of this book.

If $\pi: B(H)\rightarrow Q(H)$ is the canonical mapping onto the Calkin algebra, $A$ is a unital C*-algebra and $\phi: A\rightarrow B(H)$ is a unital completely positive map, then we say that $\phi$ is a representation modulo the compacts if $\pi\circ\phi: A\rightarrow Q(H)$ is a *-homomorphism.


Since the rank of each $\phi_n$ is finite-dimensional, finite direct sums like $\bigoplus_1^m\phi_n$ are compact (actually, they are finite-rank). So, in the quotient, $\bigoplus_1^\infty\phi_n(a)$ and $\bigoplus_m^\infty\phi_n$ are equal, and so \begin{align} \|\pi\circ\Phi(ab)-[\pi\circ\Phi(a)]\,[\pi\circ\Phi(b)]\|&=\|\pi\left(\bigoplus_1^\infty(\phi_n(ab)-\phi(a)\phi(b)\right)\|\\ &=\|\pi\left(\bigoplus_m^\infty(\phi_n(ab)-\phi(a)\phi(b))\right)\|\\ &\leq\|\bigoplus_m^\infty(\phi_n(ab)-\phi(a)\phi(b))\|\\ &=\sup_m\|\phi_n(ab)-\phi(a)\phi(b)\|\to0\\ \end{align} as $m$ goes to infinity. Thus $\pi\circ\Phi(ab)=[\pi\circ\Phi(a)]\,[\pi\circ\Phi(b)]$.

The "obvious rank-one projections" are $\displaystyle\bigoplus_{n=1}^mI_{k(n)}$, $m=1,2,\ldots$

  • $\begingroup$ Aha, I got the point. Thanks a lot~ :P $\endgroup$ – Yan kai Nov 2 '14 at 6:50

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.