Research On 3D FDTD Subgrid Technology

This paper mainly analyzes two kinds of sub grid technology, including the specific implementation methods and performance improvement program. As a simple and effective method, finite difference time domain(Finite-Difference-Time-Domain, FDTD) has a natural advantage in calculating the electromagnetic scattering characteristics of the target. It is usually used in the simulation of the target appearance with a uniform grid of fixed size; however, when faced to the large size target of the fine structure, it will inevitably lead to a large number of memory consumption and computing efficiency. Therefore, an effective treatment scheme – sub grid technology is provided at the historic moment, only applying the finer grid to the fine structure of target and coarse grid to else, which can effectively use all the resources to greatly improve the performance of calculation.Different mesh boundary processing methods bring different sub grid technologies. This paper introduces two kinds of sub grid implementation principles based on wave equation and linear interpolation, and provides the corresponding scattering and radiation examples. The numerical results present a very low reflectivity, stable program, correct and effective results.Several common FDTD modeling methods are briefly described, including the simple geometric modeling, the value point method, the triangle surface element method, and the emphasis is given to the modeling of the sub grid complex model.In order to further improve the performance of the sub grid, thus meeting the needs of large-scale parallel computing, this paper discusses the parallel implementation of sub grid based on MPI and Open MP, which greatly improves the performance of the sub grid technology.