Preconditioners based on Windowed Fourier Frames Applied to Elliptic Partial Differential Equations

TL;DR

This paper explores the use of windowed Fourier frames as preconditioners to improve the numerical solution of elliptic partial differential equations, demonstrating bounded condition numbers and effective convergence in iterative methods.

Contribution

Introduces a novel preconditioning approach based on windowed Fourier frames for elliptic PDEs, enhancing convergence properties in iterative solvers.

Findings

Condition number becomes bounded with periodic boundary conditions.

Iterative methods converge effectively despite some singular values in Dirichlet problems.

Preconditioning improves numerical stability and convergence in elliptic PDE solutions.

Abstract

We investigate the application of windowed Fourier frames (WFFs) to the numerical solution of partial differential equations, focussing on elliptic equations. The action of a partial differential operator (PDO) on a windowed plane wave is close to a multiplication, where the multiplication factor is given by the symbol of the PDO evaluated at the wave number and central position of the windowed plane wave. This can be exploited in a preconditioning method for use in iterative inversion. For domains with periodic boundary conditions we find that the condition number with the preconditioning becomes bounded and the iteration converges well. For problems with a Dirichlet boundary condition, some large and small singular values remain. However the iterative inversion still appears to converge well.