Full-Wave Analysis in SPICE via Model-Order Reduction

Fast Frequency Sweep of FEM Models via the Balanced Truncation Proper Orthogonal Decomposition

“…The surfaces where the PBCs are enforced are shown with red lines and are denoted with , and . Similar to [23], the Robin-type transmission conditions will be used to weakly enforce the PBC between master and slave boundaries, through (2) where (3) (4) are the tangential electric trace and the normalized electric current at each surface. Moreover, , is the phase difference between fields on opposite master-slave boundaries, where is the wave-vector dictated by the array scan excitation and is the displacement vector between master and slave surfaces.…”

“…Frequency interpolation schemes of -parameters or other observable quantities [3], [4] are very efficient on wideband systems, but do not provide a full EM system representation. Contrary, model order reduction (MOR) techniques [5], [6] provide a full EM system description, but are most efficient on resonant structures or problems with perturbative losses and dispersion.…”

An Efficient Framework for Analysis of Wire-Grid Shielding Structures over a Broad Frequency Range

“…Modeldriven methods are mostly aimed towards reducing the system model complexity, while preserving (to the possible extent) their input-output behavior. Representative examples of the approach are Model-Order Reduction (MOR) techniques relying on Krylov subspace methods usually combined with Lanczos or Arnoldi iteration [7], [8], and Asymptotic Waveform Evaluation (AWE) together with its extension to multipoint expansion technique referred to as Complex Frequency Hopping (CFH) [9][10][11]. In IE-MoM context, both MOR and AWE require MoM-generated system matrix to construct the reduced model or Padé-type rational approximation to the solution of the integral equation, respectively.…”

Broadband Rational Modeling via Adaptive Frequency Sampling for Full-Wave Simulation

“…The macromodels of two examples are generated to demonstrate the accuracy and efficiency of the proposed method.. The basis functions can be polynomials, such as power series or orthogonal polynomials [4], or they can also be rational functions [5]- [7]. In vector fitting, the use of simple partial fractions as basis functions in numerator and denominator, which share a common set of poles, is proposed.…”