CMV block matrices for symmetric matrix measures on the unit circle (Q2174422)

Given an \(l\times l\) positive definite matrix of measures \(\Omega\) supported on an infinite subset of the unit circle, an CMV basis of matrix Laurent orthonormal polynomials \(\{\chi_{n} (z) \}_{n\geq0}\) is defined in terms of right and left matrix orthonormal polynomials with respect to \(\Omega.\) Indeed, if \(\{\phi^{R}_{n}(z) \}_{n\geq0}\) and \(\{\phi^{L}_{n} (z) \}_{n\geq0}\) denote such sequences of orthonormal matrix polynomials, then \(\chi_{2n} (z)= z^{-n} \phi^{L, H}_{2n} (z), \chi_{2n+1} (z)= z^{-n} \phi^{R}_{2n+1}(z), n\geq 0.\) Notice that the matrix representation of the multiplication operator by \(z\) with respect to the basis is a unitary block CMV matrix. These results are presented in Section 2. If the matrix Verblunsky coefficients, i.e., the evaluation at \(z=0\) of the corresponding left and right matrix orthonormal polynomials, are Hermitian matrices with \(2\)-norm less than \(1,\) the natural counterpart in the matrix case of the well known situation in the scalar case (see Chapter 13 in [\textit{B. Simon}, Orthogonal polynomials on the unit circle. Part 2: Spectral theory. Providence, RI: American Mathematical Society (2005; Zbl 1082.42021)]), in Section 3 the relation between the block CMV matrix (in such a situation, the authors call \(\Omega\) symmetric) and the block Jacobi matrix associated with the so called Szegő transformation is stated. In particular, the entries of the block Jacobi matrix are given in terms of the Verblunsky coefficients in a similar way as in the scalar case. On the other hand,the authors generalize to the matrix case some results obtained for the scalar case in [\textit{F. Marcellán} and \textit{N. Shayanfar}, Linear Algebra Appl. 485, 305--344 (2015; Zbl 1323.42024)], when the basic matrices appearing in the factorization of the CMV matrices associated with two sequences of matrix Verblunsky coefficients are unitarily similar. In Section 4 Christoffel and Uvarov transformations of \(\Omega\) are introduced. The connection formulas for the corresponding sequences of matrix orthonormal polynomials in terms of the matrix orthonormal polynomials with respect to \(\Omega\) are deduced. Notice that Christoffel transformations in the most general framework of matrices of bi-variate measures have been studied in [\textit{J. C. García-Ardila} et al., Complex Anal. Oper. Theory 13, No. 8, 3979--4005 (2019; Zbl 1437.42038)]. The expression of their Verblunsky coefficients follows in a natural way. Finally, given two sequences of matrix Verblunsky coefficients \(\{\alpha_{n}\}_{n\geq0}\) and \(\{\hat{\alpha}_{n}\}_{n\geq0},\) an CMV block matrix associated with a sequence of matrix Verblunsky coefficients \(\{\alpha_{n,k}\}_{n\geq0},\) where \(\alpha_{n, k}= \alpha_{n} , n= rk, \, \cdots, (r+1)k-1,\) for every \(r\) when \(k\) is an even nonnegative integer number, and \(\alpha_{n, k}= \hat{\alpha}_{n} , n= rk,\, \cdots, (r+1)k-1,\) for every \(r\) when \(k\) is an odd positive integer number, is obtained in Section 5 by using very nice matrix tools.