1
$\begingroup$

Let $L$ be the left shift operator on $\ell^2(\mathbb{Z})$ with trace $\tau(T) := \langle T \delta_0, \delta_0 \rangle$.

How can I show that the Brown measure of $L$ is the uniform measure on the unit circle?

The Brown measure is defined as follows: For each $z \in \mathbb{C}$ define $\nu_{z}$ to be the spectral measure of the (self-adjoint) operator $(L - z)^*(L-z)$. Then let $$f(z) = \frac{1}{2} \int_0^\infty \log x \,d\nu_z(x).$$ Then the Brown measure is defined as $$\mu_L := \frac{1}{2\pi} \Delta f,$$ where the Laplacian is taken in the sense of distributions. Motivation for the definition can be found at Terence Tao's notes on the circular law for random matrices.

Edit: posted at math stackexchange https://math.stackexchange.com/questions/2258115/brown-measure-of-left-shift-operator

$\endgroup$
  • $\begingroup$ What's the relevance of $\tau (T)$ for this question? $\endgroup$ – Christian Remling Apr 29 '17 at 17:15
  • $\begingroup$ To address my own comment, I guess what you're saying is that $\nu_z$ is defined as the spectral measure $d\|E(x)\delta_0\|^2$ of $\delta_0$ ("the spectral measure," meaning a positive measure on $\mathbb C$, is not unambiguously defined)? $\endgroup$ – Christian Remling Apr 29 '17 at 17:17
  • 1
    $\begingroup$ The spectral measure of $L$ wrt $\delta_0$ is Lebesgue measure on $S^1$. Using this as your starting point, it should be possible to work everything out explicitly. I believe your question would have been better suited for math.stackexchange.com $\endgroup$ – Christian Remling Apr 29 '17 at 17:26
  • $\begingroup$ I was having trouble computing the spectral measure of $(L-z)^*(L-z)$, however. Is that still easy? And fair enough, I posted here because the only other references to Brown measure I could find were in research papers. I just posted in stackexchange. $\endgroup$ – keej Apr 29 '17 at 19:51
  • $\begingroup$ Actually, it might be better to do it differently. I'll answer at MSE. $\endgroup$ – Christian Remling Apr 30 '17 at 2:37
4
$\begingroup$

Since the two-sided shift is a normal operator this is not really a question about the Brown measure, but about the spectral measure, so it suffices to compute the $*$-moments $\tau(L^nL^{*m})$. For an introduction to Brown measure you might also see Chapter 11 of the book https://www.math.uni-sb.de/ag/speicher/publikationen/Mingo-Speicher.pdf

| cite | improve this answer | |
$\endgroup$

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.