A parallel implementation of the Multilevel Characteristic Basis Function Method

This paper identifies some of the issues associated with the parallel implementation of the multilevel characteristic basis function method (ML-CBFM). We describe the major characteristics of the ML-CBFM in order to explain how it can be parallelized. We present efficient schemes for the parallelization of each stage that are tailored for distributed computing. Several examples are shown to validate the performance as well as the accuracy of the results.