Synthesis of capacitive-coupled dual-behavior resonator (CCDBR) filters

A technique for the reduced-cost modeling of microwave filters is presented. Our approach exploits variable-fidelity electromagnetic (EM) simulations, and Gaussian process regression (GPR) carried out in two stages. In the first stage of the modeling process, a mapping between EM simulation filter models of low and high fidelity is established. The mapping is subsequently used in the second stage, making it possible for the final surrogate model to be constructed from training data obtained using only a fraction of the number of high-fidelity simulations normally required. As demonstrated using three examples of microstrip filters, the proposed technique allows us to reduce substantially (by up to 80%) the central processing unit (CPU) cost of the filter model setup, as compared to conventional (single-stage) GPR-the benchmark modeling method in this study. This is achieved without degrading the model generalization capability. The reliability of the two-stage modeling method is demonstrated through the successful application of the surrogates to surrogate-based filter design optimization.

show abstract

“…Consider the second-order capacitively coupled dualbehavior resonator (CCDBR) microstrip filter [24], see Figure 1.…”

Section: A Capacitively Coupled Dual-behavior Resonator Filter (Filtmentioning

confidence: 99%

Reduced-cost microwave filter modeling using a two-stage Gaussian process regression approach

Jacobs

Koziel

2014

Int J RF and Microwave Comp Aid Eng

View full text Add to dashboard Cite

show abstract

“…Consider a second-order capacitively-coupled dual-behavior resonator (CCDBR) microstrip filter [9] shown in Fig. 1 T mm.…”

Section: Verification Examples a Ccdbr Filtermentioning

confidence: 99%

Reliable low-cost co-kriging modeling of microwave devices

Koziel

Couckuyt

Dhaene

2012

2012 IEEE/MTT-S International Microwave Symposium Digest

View full text Add to dashboard Cite

Abstract-A reliable methodology for accurate modeling of microwave devices is presented. Our approach exploits cokriging which utilizes low-and high-fidelity EM simulation data and combines them into a single surrogate model. Densely sampled low-fidelity data determines a trend function which is further corrected by sparsely sampled high-fidelity simulations. Low-fidelity EM data is also enhanced by using a frequency scaling. With our method, accurate models can be obtained at a fraction of cost required by conventional approximation models that are exclusively based on high-fidelity simulations. Two cases of microstrip bandpass filters are considered. Comparisons with conventional approximation models as well as application examples are also given.Index Terms-Microwave modeling, response-surface modeling, co-kriging, electromagnetic simulation. I. INTRODUCTIONAccurate and fast models (surrogates) are indispensable in the design of microwave structures and components. Many design tasks, such as parametric optimization, statistical analysis or yield-driven optimization, require numerous evaluations of a structure of interest and the use of highfidelity electromagnetic (EM) simulations may be prohibitive because of unacceptably high computational cost.Cheap models can be obtained using response surface approximation techniques such as polynomial regression [1], radial basis functions [2], kriging [2], support vector regression [3], or artificial neural networks [4]. However, for good accuracy, these techniques require a large number of training points, which exponentially grows with the dimensionality of the design space. This high initial setup cost may be justifiable for multiple-use library models but not quite for one-time design and analysis of a specific structure.Low-cost microwave modeling can be realized using space mapping (SM) [5]. Reasonably accurate SM surrogate model can be created using a limited number of high-fidelity EM simulations by applying suitable correction to the underlying lowfidelity (or coarse) model, e.g., equivalent circuit. A drawback of SM models is that increasing the number of training points may have little effect of the model's quality [6]. Also, SM requires that the coarse model is very fast, as each evaluation of the SM surrogate requires evaluation of the underlying coarse model.In this paper, we consider models constructed using both highand low-fidelity EM simulations. Simulation of coarselydiscretized structure is less accurate but it is much faster than the

show abstract

“…Geometry of the capacitively coupled dual-behavior resonator microstrip filter [37]. to the affine mapping .…”

Section: B Capacitively Coupled Dual-behavior Resonator Filtermentioning

confidence: 99%

Space Mapping With Adaptive Response Correction for Microwave Design Optimization

Koziel

Bandler

Madsen

2009

IEEE Trans. Microwave Theory Techn.

118

View full text Add to dashboard Cite

Abstract-Output space mapping is a technique introduced to enhance the robustness of the space-mapping optimization process in case the space-mapped coarse model cannot provide sufficient matching with the fine model. The technique often works very well; however, in some cases it fails. Especially in the microwave area where the typical model response (e.g., 21 ) is a highly nonlinear function of the free parameter (e.g., frequency), the output spacemapping correction term may actually increase the mismatch between the surrogate and fine models for points other than the one at which the term was calculated, as in the surrogate model optimization process. In this paper, an adaptive response correction scheme is presented to work in conjunction with space-mapping optimization algorithms. This technique is designed to alleviate the difficulties of the standard output space mapping by adaptive adjustment of the response correction term according to the changes of the space-mapped coarse model response. Examples indicate the robustness of our approach.

show abstract

Synthesis of capacitive-coupled dual-behavior resonator (CCDBR) filters

Cited by 63 publications

References 13 publications

Reduced-cost microwave filter modeling using a two-stage Gaussian process regression approach

Reduced-cost microwave filter modeling using a two-stage Gaussian process regression approach

Reliable low-cost co-kriging modeling of microwave devices

Space Mapping With Adaptive Response Correction for Microwave Design Optimization

Contact Info

Product

Resources

About