The commutant lifting approach to interpolation problems (Q1188738)

Let \(H^{\infty}\) be the Hardy space of bounded analytic functions in the unit disc \(D\) of the complex plane \(\mathbb{C}\) and let \(H^{\infty}_ 1\) be the unit ball of \(H^{\infty}\). At the beginning of the century C. Carathéodory considered the following interpolation problem: (P.1). Given a string \((a_0,a_1,\ldots,a_{n-1})\) in \(\mathbb{C}^n\) find necessary and sufficient conditions for the existence of a function \(f\in H^{\infty}_1\) such that \[ f(z)=a_0+a_1z+\cdots+a_{n-1}z^{n-1}+O(z^n). \] The string \((a_0,a_1,\ldots,a_{n-1})\) is called the data of the Carathéodory problem (of length \(n)\) for the function \(f\in H^{\infty}_1\). At the same time I. Schur posed and solved the problem of giving an explicit description of all solutions (if any) of the Carathéodory problem. Obviously that, by the maximum principle, the Carathéodory interpolation problem has no solution for \(| a_0| >1\) and has the unique solution \(f(z)=a_0\) for \(| a_0| =1\). In a now classical paper \textit{D. Sarason} [Trans. Am. Math. Soc. 127, 179--203 (1967; Zbl 0145.39303)], realized that the interpolation problems are connected with the problem of lifting two commuting contractions. Shortly after that, \textit{B. Sz.-Nagy} and \textit{C. Foiaş} [C. R. Acad. Sci. Paris, Sér. A 266, 493--495 (1968; Zbl 0173.16503)], obtained a purely geometric extension of Sarason's result: Operators intertwining restrictions of co-isometries can be extended, by preserving their norm, to operators intertwining co-isometries. It turned out over the next years that this deep result had many consequences for interpolation, \(H^{\infty}\) approximation, the Toeplitz Corona Theorem. The aim of this book is to emphasize the relevance of the commutant lifting theorem for these branches of complex and functional analysis and also to present applications to some engineering problems, especially in seismic oil prospecting. Curiously that, geophysicists working in this field independently arrived at the Carathéodory interpolation problem and rediscovered the Schur algorithm in 1960's. The first three chapters of the book, I. Analysis of the Carathéodory interpolation problem and II. Analysis of the Carathéodory interpolation problem for positive-real functions, III. Schur numbers, geophysics and inverse scattering problems, are devoted to a rather complete presentation of classical Carathéodory interpolation problem and of various algorithms for solving the Schur problem - the original Schur's algorithm based on the Möbius transform to peel off the data and of other faster algorithms, including the layer peeling algorithm used in geophysics and Levinson algorithm from filtering theory. Using elementary technique from operator theory (in \({\mathbb{C}}^ n)\) one shows that the Carathéodory interpolation problem is solvable if and only if certain matrix is a contraction. These chapters motivate many of the developments in this monograph and the authors continue to return to the Carathéodory interpolation problem as the standard test case. The study of Hilbert space case begins with Chapter IV. Contractive explanations on Euclidean and Hilbert space. This one and the following two - Chapter V. Contractive one step intertwining liftings and Chapter VI. Isometric and unitary dilations, contains some preliminary material needed for the central result of the book, exposed in Chapter VII. The commutant lifting theorem. The authors present four proofs to this theorem each of them illuminating different features of this very important and deep result. The first one is a modification of the original proof of B. Sz.-Nagy and C. Foiaş, the second is used to discuss uniqueness problem, the third establishes the connection to Ando's dilation theorem and the fourth is based on Arocena's coupling of contractions. Chapter VIII. Geometric applications of the commutant lifting theorem is concerned with operator version of the classical interpolation problems of Carathéodory, Fejér, Nevanlinna-Pick and with Hankel matrix interpolation problem. The commutant lifting theorem is used also to prove some Corona theorems for analytic Toeplitz operators. One of the problems considered in Chapter IX. \(H^{\infty}\) optimization and functional models, is the Nehari optimization problem: For \(Q\in L^{\infty}(E,E')\), \((E,E'\) Hilbert spaces), find \(\mu = \operatorname{dist}(Q,K^{\infty}_0(E,E'))\), where \(L^{\infty}(E,E')\) denotes the space of all strongly measurable a.e. bounded functions on \([0,2\pi]\) with values in \(L(E,E')\) and \(K^{\infty}_0(E,E')\) its subspace formed of all functions in \(L^{\infty}(E,E')\) whose Fourier coefficients with positive indices are all zero. The problem has always a solution which is a Hankel operator from \(H^2(E)\) to \(H^2(E')\). In Chapter X, Some classical interpolation problems, it is shown that there exists a natural connection between Hermite-Fejér interpolation theory and contractive Hankel operators with rational symbols giving also a solution to Nehari \(H^{\infty}\)-optimization problem. In Chapter XI, Interpolation as a momentum problem, the interpolation problems are solved by using a general momentum theorem proved by M. Rosenblum and J. Rovnyak in 1980. In Chapter XII, Numerical algorithms for \(H^{\infty}\)-optimization in control theory, one shows how the \(H^{\infty}\)-optimization problems considered in Chapter IX naturally arise in robust control theory. The aim of Chapter XIII, Inverse scattering algorithms for the commutant lifting theorem, is to obtain a complete characterization of all contractive intertwining liftings in the commutant lifting theorem. In Chapter XIV, The Schur representation, the Redheffer products from network theory are used to obtain Schur type representations in the commutant lifting theorem. The fundamental result of Chapter XV. A geometric approach to positive definite sequence, is the Naimark dilation theorem for positive definite sequences of operators which is then applied to solve the positive definite Carathéodory interpolation problem. Chapter XVI. Positive definite block matrices, is a positive matrix version of Chapter IV. Complete characterizations of all 2 by 2 and 3 by 3 positive block matrices are given. In the last chapter, Chapter XVII, A physical basis for the layered medium model, it is sketched a physical framework for Chapter III. This concerns the study of vertically moving horizontal primary waves through layered medium. The book is very well and clearly written and contains a wealth of results, some of the basic results in the field belonging to the authors. The authors tried to make the book as self-contained as possible. A graph indicating the dependence between chapters will help the reader to reach quickly the matters interesting him. By presenting deep results of interpolation theory within the framework of Hilbert space operator theory and their nontrivial and important applications in geophysics, the book will be of great interest both for those interested in theoretical aspects as well as for those interested mainly in engineering applications.