Solution of a tridiagonal operator equation

DOI10.1016/J.LAA.2005.10.014MaRDI QIDQ819790zbMATH OpenOpenAlexFDO

Authors R. Balasubramanian, S. H. Kulkarni, R. Radha

Publication date 29 March 2006

Published in Linear Algebra and its Applications (Search for Journal in Brave)

Full work available at URL https://doi.org/10.1016/j.laa.2005.10.014

zbMATH Keywords

determinantdiagonal dominanceGerschgorin disctridiagonal matrixtridiagonal operator

Mathematics Subject Classification ID

Norms of matrices, numerical range, applications of functional analysis to matrix theory (15A60) Equations and inequalities involving linear operators, with vector unknowns (47A50) Linear operators on special spaces (weighted shifts, operators on sequence spaces, etc.) (47B37) Numerical solutions to equations with linear operators (65J10)

Review text

Let \(T\) be the tridiagonal operator \[ T= \left[\begin{matrix} d_1 & c_1\\ u_2 & d_2 & c_2\\ & u_3 & d_3 &\ddots\\ &&\ddots &\ddots\end{matrix}\right] \] on \(\ell^2\), where \(\{c_n\}\), \(\{d_n\}\) and \(\{u_n\}\) are bounded sequences of complex numbers. For \(n\geq 1\), let \(T_n\) be the \(n\)-by-\(n\) leading principal submatrix of \(T\). This paper studies the operator equation \(Tx= y\) through its truncated equations \(T_n x^n= y_n\), where \(y_n\) is the vector of the first \(n\) components of \(y\). It is shown that if \(T_n\) is invertible for all \(n\) and \(\{\| T^{-1}_n e_n\|\}\) is bounded, where \(e_n\) denotes the \(n\)-vector \([0,\dots, 0,1]^t\), then any solution \(x\) of \(Tx= y\) is the norm limit of \([x^n,0]^t\), where \(x^n\) is such that \(T_n x^n= y_n\) and, in particular, \(T\) is one-to-one. Moreover, if the \(c_n\)'s and \(u_n\)'s are all nonzero and \(\{\| {T^\ast}^{-1}_n e_n\|\}\) is also bounded, then \(T\) is invertible. The authors then provide some verifiable conditions on the entries of \(T\) for the boundedness of \(\{\| T^{-1}_n e_n\|\}\) and \(\{\| {T^\ast}^{-1}_n e_n\|\}\). One such condition is that \(c_n= u_n= 1\) and for some \(\varepsilon> 0\), \((|d_n d_{n+1}|- 2)(|d_{n-1} d_n|- 2)\geq 4+\varepsilon\) for all \(n\).

Recommendations

• On the explicit inverse and conditioning of a tridiagonal matrix
• Bounds for the solutions of a class of tridiagonal linear systems
• Perron-frobenius-type theorems for tridiagonal operators
• Solution of an infinite band matrix equation
• scientific article; zbMATH DE number 3970940

Cites work

• \(C^*\)-algebras and numerical linear algebra
• scientific article; zbMATH DE number 3854346 (Why is no real title available?)
• scientific article; zbMATH DE number 43732 (Why is no real title available?)
• scientific article; zbMATH DE number 49612 (Why is no real title available?)
• Non-invertibility of certain almost Mathieu operators
• Spectral and structural analysis of high precision finite difference matrices for elliptic operators
• The role of C*-algebras in infinite dimensional numerical linear algebra

Cited in

(6)

• On the existence of solutions for the Frenkel-Kontorova models on quasi-crystals
• Solution of an infinite band matrix equation
• Invertibility of a tridiagonal operator with an application to a non-uniform sampling problem
• Enumeration of smooth inversion sequences and proof of a recent related conjecture
• Frames in generalized Fock spaces
• Local reconstruction of a function from a non-uniform sampled data

This page was built for publication: Solution of a tridiagonal operator equation

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q819790)