Convolution operators and factorization of almost periodic matrix functions (Q700450): Difference between revisions

The book systematically describes the factorization problem for almost periodic (AP) and semi-almost periodic (SAP) functions. Such a problem arises in the study of many classes of integral operators (for instance, convolution operators or Toeplitz operators with almost periodic symbols, convolution operators on finite intervals, etc.). The factorization problem for continuous or piecewise continuous matrix functions (and, in general, for measurable matrix functions) was thoroughly studied by now, and is presented in many reviews and books [see, for example, \textit{G. S. Litvinchuk} and \textit{I. M. Spitkovsky}, ``Factorization of measurable matrix functions''. Basel-Boston: Birkhäuser (1987; Zbl 0651.47010)]. Although the factorization of AP matrix-functions has already been studied for more than 20 years, it is done here in a monograph for the first time. The book presents advances and developments of this theory and its applications including all modern results and directions. The book contains 23 chapters. First, the authors give a brief contemporary analysis of the theory of several classes of scalar and matrix convolution type operators with continuous and piecewise continuous symbols, and describe the problem of factorization for the analogous matrix-functions. The primary attention of the authors is certainly directed to more complicated case of oscillating symbols which leads to the factorization of almost periodic or semi-almost periodic matrix functions. For example, such symbols arise in the study of the convolution operator with symbol \(f\) on a finite interval. The solvability problem for such operators on an interval \((0, \lambda)\) is related to the factorization of its oscillating matrix-symbol \(G^\lambda_f (x)\) on the half-axis. Here \(G^\lambda_f (x)\) is a triangular \(2\times 2\) matrix-function of the form \[ \left(\begin{matrix} e^{i\lambda x} & 0 \\ f(x) & e^{-i\lambda x} \end{matrix}\right). \] Some new phenomena caused by SAP symbols are studied. The authors study several forms of the factorization of AP matrix-functions. One of them is the representation of \(G^\lambda_f (x)\) with \(f \in \text{AP}\) in the form \(G^\lambda_f (x)=G_-(x)\;\text{diag} (e^{i\mu_1 x}, e^{i\mu_2 x}) G_+ (x)\) where \(G_\pm (x)\) are invertible AP matrix-functions. The Bohr-Fourier spectrum of \(G_\pm^{\pm 1} (x)\) is in \([0, \infty),\) \((-\infty, 0]\), respectively, and the real numbers \(\mu_1, \mu_2\) stand for almost periodic indices of \(G^\lambda_f (x)\). In general, a matrix function \(a(x)\in \text{GAP}_{N\times N}\) has a right factorization if \(a(x)=a_- (x)\operatorname {diag}(e^{i\lambda_1 x},e^{i\lambda_2 x},\dots , e^{i\lambda_N x})a_+(x)\) for all \(x\in \mathbb{R}\) with \(a_\pm(x)\in \text{GAP}^\pm_{N\times N}\) and \(\lambda_j \in\mathbb{R}\). Here \(AP^\pm=\text{alg}_{L^\infty(R)}e^{\pm i\lambda x}\), \(\lambda \geq 0\},\) and GAP consists of invertible elements of AP. The factorization \(a(x)=a_+ (x)\operatorname {diag}(e^{i\lambda_1 x},e^{i\lambda_2 x},\dots , e^{i\lambda_N x})a_-(x)\) is called the left one. In the present monograph, the authors consider the problem of AP factorization for many classes of matrix-functions and give applications to many classes of operators. The connection of the problems under consideration with the corona problem for AP and APW is also discussed. The \(C^\ast\)-algebra \(SAP\) is defined as \(\text{SAP}^\pm=\text{alg}_{L^\infty(\mathbb{R})} (\text{AP}, C(\overline{\mathbb{R}}))\) and there exist representatives \(a_{\ell}, a_r \in \text{AP}\) such that the maps \(a\mapsto a_{\ell}, a\mapsto a_r\) are \(C^\ast\)-algebra homomorphisms of SAP onto AP. For \(a\in \text{AP}\) let \(M(a)=\lim_{|I|\to \infty} \frac{1}{|I|}\int_{I} a(x) dx\) (the limit does not depend on the choice of the set of intervals \(I\subset \mathbb{R}\) such that \(|I|\to \infty\)). If \(a\in \text{SAP}_{N\times N}\), the authors define a pair of pre-factorizations of its AP representatives \(a_{\ell}\), \(a_r \in \text{AP}\) in the form \[ a_{\ell}=a^-_{\ell} \begin{pmatrix} v_{\ell} & 0 \\ I & 0 \end{pmatrix} a^+_{\ell},\qquad a_r=a_r^- \begin{pmatrix} v_r & 0 \\ I & 0 \end{pmatrix} a_r^+ \] where \(a^\pm_{\ell}, a^\pm_r \in \text{GAP}^\pm_{N\times N}\) and \(v_{\ell}, v_r\) are unitary valued functions such that \[ \begin{aligned} v_\ell &\in \text{APW}^-_{m_{\ell}\times m_{\ell}}\cup \text{APW}^+_{m_{\ell}\times m_{\ell}}, \\ M(v_{\ell})&=0, \\ v_r &\in \text{APW}^-_{m_r\times m_r}\cup \text{APW}^+_{m_r\times m_r},\\ M(v_r)&=0 \text{ and} \\ m_{\ell},m_r &\in \{0,1, \dots , N\}.\end{aligned} \] Here \(\text{APW} \subset \text{AP}\) and \(\text{APW} =\{a: a=\sum_{j}a_j e^{i\lambda_j x}\), \(\{a_j\}\in \ell_1\}\). Some cases when the pair of pre-factorizations exists are given. The connection between problems with SAP symbols and PC symbols is also considered. The book should be very interesting for specialists in the field of integral operators and operator theory, and also may be recommended to researchers and postgraduate students working in pure and applied mathematics. It contains many interesting examples and very remarkable geometric illustrations. Besides the above specified, there is a wide and up to date bibliography which reflects the ultimate achievements in this field. Some unsolved problems and new interesting questions are also discussed. Contents: Preface. Chapter 1. Convolution operators and their symbols. Chapter 2. Introduction to scalar Wiener-Hopf operators. Chapter 3. Scalar Wiener-Hopf operators with SAP symbols. Chapter 4. Some phenomena caused by SAP symbols. Chapter 5. Introduction to matrix Wiener-Hopf operators. Chapter 6. Factorization of matrix functions. Chapter 7. Bohr cementification. Chapter 8. Existence and uniqueness of AP factorization. Chapter 9. Matrix Wiener-Hopf operators with APW symbols. Chapter 10. Matrix Wiener-Hopf operators with SAPW symbols. Chapter 11. Left versus right Winer-Hopf factorization. Chapter 12. Corona theorems. Chapter 13. The Portuguese transformation. Chapter 14. Some concrete factorizations. Chapter 15. Scalar transforms. Chapter 16. Toeplitz operators. Chapter 17. Zero-order pseudo-differential operators. Chapter 18. Toeplitz operators with SAP symbols on Hardy spaces. Chapter 19. Wiener-Hopf operators with SAP symbols on Lebesgue spaces. Chapter 20. Hankel operators on Besicovitch spaces. Chapter 21. Generalized AP factorization. Chapter 22. Canonical Wiener-Hopf factorization via Corona problems. Chapter 23. Canonical APW factorization via Corona problems. Bibliography. Index.