Minimal Rational Interpolation and its Application in Fast Broadband Simulation

Broad bandwidth simulation in frequency domain benefits from fast frequency sweep, and frequency-domain electromagnetic solvers are usually combined with asymptotic wave evaluation or interpolation techniques. In this work, we introduce the minimal rational interpolation from the perspective of linear algebra and utilize it in fast broadband simulation. An efficient approach is formed by combining adaptive frequency sampling and rational interpolating. Approximate models of the quantities of interests are updated iteratively until the the broadband spectral dynamics are captured. Thanks to the minimal rational interpolation, the approach is fully automatic and does not require pre-estimates of the orders of interpolating functions in each step of the iterations. The obtained functions always have low complexities, hence no further model order reduction is needed. Broadband simulations on admittance of a multi-port interconnects and backward scattering of a cube are given as numerical examples. Results show that the approach can significantly improve the computational efficiency.