Error analysis for the multilevel fast multipole algorithm

The matrix-vector multiplication encountered in the iterative solution of scattering problems can be performed in O(N) operations by using a multilevel past multipole algorithm (MLFMA). The estimates for the errors introduced by MLFMA is presented along with the details of the algorithm. An analysis of the errors in the FMM algorithm for the monopole and dipole terms is given and the analysis is extended to higher order multipole terms, and integration and interpolation errors are also included.link_to_subscribed_fulltex