The Petrov-Galerkin method for second kind integral equations. II: Multiwavelet schemes (Q1373116)

[For part I see \textit{Z. Chen} and \textit{Y. Xu}, SIAM J. Numer. Anal. 35, No. 1, 406-434 (1998; Zbl 0911.65143).] The purpose of this paper is to study wavelet based Petrov-Galerkin methods for Fredholm integral equations of the second kind and especially to investigate matrix compression resulting from the use of discontinuous orthogonal multiwavelets on the interval \([0,1].\) The obtained results show that the considered method for integral equations yields linear systems having numerically sparse coefficient matrices whose condition numbers are bounded. The authors develop a truncation strategy that leads to linear systems having sparse coefficient matrices and thus yields fast algorithms for solving this class of integral equations. Moreover, these schemes produce approximate solutions having almost optimal order of convergence. The computational complexity of these algorithms is analysed.