The method of fundamental solution (MFS) has been known as a simple and effective boundary meshless method. However, the MFS generates dense square coefficient matrix and thus requires a large amount of computational time and memory storage for solving large-scale problems. The fast multipole method (FMM) is a technique that reduces computational operations and storage requirements for solving such dense matrix equations. This study makes the first attempt to apply the FMM to the MFS

more »

... of acoustic problems, where the operations are reduced to O(N log N ) while O(N 2 ) operations are required for the traditional MFS using the standard iterative methods. Numerical examples with up to 100,000 points are successfully tested on a desktop personal computer. Our results clearly demonstrate efficiency and accuracy of the fast multipole MFS for solving large-scale Helmholtz-type problems.