An efficient algorithm for large finite periodic arrays of microstrip patches using sub-entire domain basis functions

A variety of computational techniques have been developed for the analysis (e.g., radiation, scattering and radar cross section (RCS) calculations) of large-scale finite periodic microstrip patch arrays due to their wide range of practical applications. Among the full-wave analysis techniques, the Method of Moments (MoM) is a popular choice because it discretizes only the microstrip patches, leading to a relatively small number of unknowns compared to other methods [1]. However, the conventional MoM requires O(N
²
) memory and O(N
³
) computational complexity for the direct matrix inversion solvers. These requirements make the conventional MoM inefficient for analysis of large-scale finite periodic microstrip patch arrays. To overcome this drawback, two major categories of techniques have been developed to improve the efficiency of MoM-based full-wave solvers: (a) ones that accelerate calculation of the matrix-vector product within the iterative solver via a fast Fourier transform (FFT)-based algorithm (e.g., the Adaptive Integral Equation (AIM) method [2] and the precorrected-FFT method [3]), and (b) ones that reduce the number of unknowns by employing physically based entire-domain or sub-entire-domain basis functions or high-order sub-domain basis functions (e.g., the macro-basis function method [4], the synthetic basis function method [5] and the characteristic basis function method [6]).