Constructive methods of Wiener-Hopf factorization (Q1088143): Difference between revisions

Given a simple contour \(\Gamma\) in the complex plane and a continuous \(n\times n\) matrix function W(\(\lambda)\) on \(\Gamma\), a representation \[ (1)\quad W(\lambda)=W_-(\lambda)D(\lambda)W_+(\lambda) \] is called a (right) Wiener-Hopf factorization of \(W(\lambda)\) with respect to \(\Gamma\) if \(W_-(\lambda)\) is analytic and invertible outside \(\Gamma\) (including infinity), \(W_+(\lambda)\) is analytic and invertible inside \(\Gamma\), and \(D(\lambda)\) is a diagonal matrix with the functions \((\lambda -\lambda_ 0)^{\alpha_ 1},...,(\lambda -\lambda_ 0)^{\alpha_ n}\) on the main diagonal (here \(\lambda_ 0\) is certain point inside \(\Gamma\) and \(\alpha_ 1,...,\alpha_ n\) are integers). Wiener-Hopf factorization and its numerous variations and extensions (especially for the case when \(\Gamma\) is the real line) appear in different mathematical fields, theoretical and applied. The book is a collection of 9 research papers The unified theme is the study in terms of the matrices A, B, and C, of Wiener-Hopf factorizations of rational matrix functions represented in the form \[ (2)\quad W(\lambda)=I_ n+C(\lambda I_ m-A)^{-1_ B}, \] where A,B,C are constant matrices, as well as extensions and generalizations of this approach. The impetus to this study, and subsequent applications come from the linear systems theory. The first paper ''Left versus right canonical Wiener-Hopf factorization'' (pp. 1-37), by \textit{J. A. Ball} and \textit{A. C. M. Ran}, concerns connections between left (i.e. the order of factors in (1) is interchanged) and right canonical (that is, all the integers \(\alpha_ 1,...,\alpha_ n\) are zeros) Wiener-Hopf factorization. The second paper ''Winer-Hopf equations with symbols analytic in a strip'' (pp. 39-74), by \textit{H. Bart, I. Gohberg} and \textit{M. A. Kaashoek}, extends the previously known results which describe the canonical Wiener- Hopf factorization on the real line in terms of invariant subspaces, to the class of functions W(\(\lambda)\) that are Fourier transforms of matrix functions with entries in \(e^{w| t|}L_ 1(-\infty,\infty)\), where w is some negative constant. The third paper ''On Toeplitz and Wiener-Hopf operators with contourwise rational matrix and operator symbols'' (pp. 75-126), by \textit{I. Gohberg, M. A. Kaashoek, L. Lerer} and \textit{L. Rodman}, generalizes the same previously known results to the case when \(\Gamma\) is a composite contour and on each connected component of \(\Gamma\) W(\(\lambda)\) has the form (2) with A, B, C depending on the connected components. The fourth paper ''Canonical pseudo spectral factorization and Wiener-Hopf integral equations'' (pp. 127-156), by \textit{L. Roozemond}, considers the case when \(\Gamma\) is the real line and det W(\(\lambda)\) has zeros on \(\Gamma\). The fifth paper ''Minimal factorization of integral operators and cascade decompositions of systems'' (pp. 157-220), by \textit{I. Gohberg} and \textit{M. A. Kaashoek}, extends the concepts of canonical Wiener-Hopf and minimal factorization into the direction of integral operators. This leads to a cascade decomposition theory for time-varying linear systems with well- posed boundary conditions. The sixth paper ''Explicit Wiener-Hopf factorization and realization'' (pp. 235-316), by \textit{H. Bart, I. Gohberg} and \textit{M. A. Kaashoek}, gives the construction of explicit formulas for \(W_-(\lambda)\), \(W_+(\lambda)\) and for the integers \(\alpha_ 1,...,\alpha_ n\) in terms of A, B, C. The seventh paper ''Invariants for Wiener-Hopf equivalence of analytic operators functions'' (pp. 317-356), by \textit{H. Bart, I. Gohberg} and \textit{M. A. Kaashoek}, concerns operator functions that are analytic on the real line and at infinity. The main result gives a necessary condition for Wiener-Hopf equivalence of two such operator functions. In the eighth paper ''Multiplication by diagonals and reduction to canonical factorization'' (pp. 357-372), by \textit{H. Bart, I. Gohberg} and \textit{M. A. Kaashoek}, the operation of multiplication by diagonal rational matrix functions is analyzed in geometric terms, and as a corollary a method is obtained to reduce rational matrix functions to functions that admit canonical factorization. The last paper ''Symmetric Wiener-Hopf factorization of self-adjoint rational matrix functions and realization'' (pp. 373-409), by \textit{M. A. Kaashoek} and \textit{A. C. M. Ran}, describes explicitly the Wiener-Hopf factorization on the real line for the case when \(W(\lambda)\) has selfadjoint values on the real line and the factorization is also required to be of symmetric type.