A Neural Network method for the multilayered shielded Green's function calculation

Multilayered boxed Green's Functions are needed in Integral Equation (IE) approaches, which combined with the method of moments (MoM) algorithm allows for the efficient analysis of printed shielded circuits. This analysis is computational time consuming, specially due to the slow convergence of the relevant boxed Green's functions. Neural Networks as Radial Basis Function Networks can effectively approximate the multilayered media boxed Green's functions, allowing a faster performance in the analysis of practical shielded MMIC devices.