A parallel FDTD algorithm based on domain decomposition method using the MPI library

Abstract

We present parallel algorithm for FDTD method. The algorithm is based on DDM (domain decomposition method). We use the MPI (message passing interface) library to implement the parallel DD-FDTD method. The original domain is decomposed into several subdomains according to the features of the problem, which are treated as the processes to computation nodes. The FDTD computation is carried out independently in local meshes in each process. The data are exchanged by communication between adjacent subdomains or with overlapped meshes and the interpolation scheme is applied to the data exchanged. The example of its application is given.