Convergence of joint moments for independent random patterned matrices (Q717892)

One of the fundamental links between random matrix theory and free probability is provided by the fact that, as the size of a matrix goes to infinity, families of independent Wigner matrices tend to free families of semicircular variables. In the article under review, the authors ask what happens if one replaces Wigner matrices by patterned matrices. More precisely, they study the joint convergence of a finite family of independent symmetric matrices with identical pattern. It turns out that for symmetric circulant and reverse circulant matrices, the limits are classical independent normal and half independent Rayleigh, respectively. The limits of Hankel and Toeplitz matrices are neither (half) independent nor free.