Wavelet Galerkin Algorithms for Boundary Integral Equations

From MaRDI portal





This very interesting and informative paper is directed to one of the important problems in applying wavelets to boundary integral equations: How can the asymptotic estimates for optimal compression, which means a priori neglecting of specific matrix entries, be transfered to realizable at least nearly optimal, i.e. optimal with an additional log-term and with moderate constansts, algorithms for relevant problem classes. A lot of results are summarized from earlier partly not even published papers or preprints. Thus, the authors give implicitely some kind of overview about the state of the art in applying multiwavelets to boundary integral equations. In particular, the reader is very carefully referred to the papers containing the detailed analysis or arguments. One of the central issues is the presentation of various numerical experiments figuring out the influence of various parameters. Additionally new estimates for the stiffness matrix entries assuming analytical kernel functions are given. For more details on the paper let me cite the very clear abstract by the authors: ``The implementation of a fast wavelet-based Galerkin discretization of second kind integral equations on piecewise smooth surfaces \(\Gamma\subset \mathbb R^3\) is described. It allows meshes consisting of triangles as well as quadrilaterals. The algorithm generates a sparse, approximate stiffness matrix with \({\mathcal N}= O(N(\log N)^2)\) nonvanishing entires in \(O(N(\log N)^4)\) operations, where \(N\) is the number of degrees of freedom on the boundary, while essentially retaining the asymptotic convergence rate of the full Galerkin scheme. A new proof of the matrix-compression estimates is given based on derivative-free kernel estimates. The condition number of the sparse stiffness matrices is bounded independently of the meshwidth. The data structure containing the compressed stiffness matrix is described in detail: it requires \(O({\mathcal N})\) memory and can be set up in \(O({\mathcal N})\) operations. Numerical experiments show that the asymptotic performance estimates apply for moderate \(N\). Problems with \(N=10^6\) degrees of freedom were computed in core on a workstation. The impact of various parameters in the compression scheme on the performance and the accuracy of the algorithm is studied.




Cited in
(75)








This page was built for publication: Wavelet Galerkin Algorithms for Boundary Integral Equations

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q4702361)