On a sparse representation of an \(n\)-dimensional Laplacian in wavelet coordinates (Q2634307)

The authors construct a well-conditioned wavelet basis with homogeneous boundary conditions on the unit interval \([0,1]\). This wavelet basis is based on Hermite cubic splines so that both the mass matrix and the stiffness matrix corresponding to the one-dimensional Poisson equation are sparse, i.e., the number of nonzero elements in each column is bounded independently of the matrix size. Condition numbers of the one-dimensional stiffness matrices, resp. mass matrices are presented for different decomposition levels. The proposed wavelet basis is well conditioned for low decomposition levels.