Two-dimensional ARMA for FDTD Radiation Pattern Analysis

Abstract

The finite-difference time-domain (FDTD) method has been widely employed for analyzing various electromagnetic problems, including low-frequency problems. It provides accurate results in most cases, however, relatively long calculation time is required for slow convergence problems, such as array antennas. For this reason, the autoregressive moving average (ARMA) model has been applied to FDTD radiation pattern analysis because the radiation patterns calculation requires multi-dimensional (time and space) data. In this paper, we applied two-dimensional ARMA to FDTD radiation pattern analysis to reduce the calculation time. We confirm effectiveness of the method by calculating the radiation pattern of Yagi-Uda antenna.