An Implicit Space Mapping Technique for Component Modeling

Abstract

Abstract-We present an implicit Space Mapping (ISM) technique for modeling microwave and RF components, the first attempt to employ ISM-based surrogates specifically for modeling. The proposed technique calibrates the surrogate by mapping certain preassigned parameters.This surrogate employs linearly interpolated preassigned parameters. Using a three-section transformer, we demonstrate that our technique is simple to apply within Agilent ADS. We show that we can easily enhance an HTS filter empirical model to the accuracy of an electromagnetic model, Sonnet's em.

A New Multi-Dimensional RF and Microwave Modeling Algorithm Based on Rational Interpolants and Hybrid Mapping

“…This phase is motivated by the space-mapping concept [6] and the authors' experience in RF/microwave modeling [4]. The objective of Phase 2 is to develop a set of rational sub-model which represent the given RF/microwave behavior at a new x = Xnew i.e.…”

“…Along a dimension, identified as the most non-linear (often the frequency axis), an enhanced adaptive sampling algorithm based on rational interpolants is used to create a sub-model. Inspired by space-mapping [6], the validated sub-model is mapped along the other, less non-linear dimensions using a novel hybrid mapping scheme. Consequently, the overall multi-dimensional model can be generated.…”

Implicit space mapping with adaptive selection of preassigned parameters

“…Another advantage of implicit space mapping is that, unlike input space mapping [1,4], it does not affect the domain of the surrogate model, which may be important in the case of constrained optimisation. Implicit space mapping has been demonstrated to work both in the design [3,7,35,36] and modelling [37,38] arenas.…”

Tuning-aided implicit space mapping

“…Implicit space mapping (ISM) has been demonstrated to work both in the design [3] and modeling [4] arenas. ISM [3,4] optimization techniques exploit available preassigned parameters in the coarse model.…”

“…ISM [3,4] optimization techniques exploit available preassigned parameters in the coarse model. The preassigned parameters are a set of parameters that are normally fixed during the design optimization process, such as dielectric constant and substrate height.…”