Title: Modelling and Design of a Microstrip Band-Pass Filter Using Space Mapping Techniques
ArXiv ID: 1004.4594
Date: 2010-04-27
Authors: Researchers from original ArXiv paper
📝 Abstract
Determination of design parameters based on electromagnetic simulations of microwave circuits is an iterative and often time-consuming procedure. Space mapping is a powerful technique to optimize such complex models by efficiently substituting accurate but expensive electromagnetic models, fine models, with fast and approximate models, coarse models. In this paper, we apply two space mapping, an explicit space mapping as well as an implicit and response residual space mapping, techniques to a case study application, a microstrip band-pass filter. First, we model the case study application and optimize its design parameters, using explicit space mapping modelling approach. Then, we use implicit and response residual space mapping approach to optimize the filter's design parameters. Finally, the performance of each design methods is evaluated. It is shown that the use of above-mentioned techniques leads to achieving satisfactory design solutions with a minimum number of computationally expensive fine model evaluations.
💡 Deep Analysis
Deep Dive into Modelling and Design of a Microstrip Band-Pass Filter Using Space Mapping Techniques.
Determination of design parameters based on electromagnetic simulations of microwave circuits is an iterative and often time-consuming procedure. Space mapping is a powerful technique to optimize such complex models by efficiently substituting accurate but expensive electromagnetic models, fine models, with fast and approximate models, coarse models. In this paper, we apply two space mapping, an explicit space mapping as well as an implicit and response residual space mapping, techniques to a case study application, a microstrip band-pass filter. First, we model the case study application and optimize its design parameters, using explicit space mapping modelling approach. Then, we use implicit and response residual space mapping approach to optimize the filter’s design parameters. Finally, the performance of each design methods is evaluated. It is shown that the use of above-mentioned techniques leads to achieving satisfactory design solutions with a minimum number of computationally e
📄 Full Content
JOURNAL OF TELECOMMUNICATIONS, VOLUME 2, ISSUE 1, APRIL 2010
Modelling and Design of a Microstrip Band-
Pass Filter Using Space Mapping Techniques
Saeed Tavakoli, Mahdieh Zeinadini, Shahram Mohanna
Abstract—Determination of design parameters based on electromagnetic simulations of microwave circuits is an iterative and
often time-consuming procedure. Space mapping is a powerful technique to optimize such complex models by efficiently
substituting accurate but expensive electromagnetic models, fine models, with fast and approximate models, coarse models. In
this paper, we apply two space mapping, an explicit space mapping as well as an implicit and response residual space
mapping, techniques to a case study application, a microstrip band-pass filter. First, we model the case study application and
optimize its design parameters, using explicit space mapping modelling approach. Then, we use implicit and response residual
space mapping approach to optimize the filter’s design parameters. Finally, the performance of each design methods is
evaluated. It is shown that the use of above-mentioned techniques leads to achieving satisfactory design solutions with a
minimum number of computationally expensive fine model evaluations.
Index Terms—Explicit space mapping, implicit and response residual space mapping, microstrip band-pass filter, modeling and
design, surrogate model.
—————————— ——————————
1 INTRODUCTION
ONSIDERING the development of computer-aided
design techniques, optimization plays a vital role in
modelling and design of microwave circuits. A typi-
cal design problem is to choose the design parameters to
get the desired response. Space mapping (SM) approach,
introduced by Bandler et al. [1], is a powerful technique
to optimize complex models. It substitutes efficiently ex-
pensive electromagnetic models, fine models, with fast
and approximate models, coarse models. To obtain the
optimal design for the fine model, the SM establishes a
mapping between parameters of the two models iterative-
ly [1, 2]. SM techniques can be classified to original or
explicit SM [3] and implicit SM (ISM) [4] methods. Both
methods use an iterative approach to update the mapping
and predict new design parameters.
Explicit space mapping modelling approach is based
on setting up a surrogate model. SM-based surrogate
models involve only certain combinations of input and
output mappings. The input mapping is an explicit map-
ping between design parameters of the coarse and fine
models. It is aimed to match SM-based surrogate and fine
models in a region of interest.As evaluation of fine mod-
els is expensive, surrogate models, which should be fast,
accurate and valid in a wide range of parameters, should
be constructed using only a few fine model evaluations.
In other words, SM-based surrogate models should use a
small amount of data from fine models to extract the in-
put and output’s mapping parameters. Having the space
mapping parameters established, the evaluation of SM-
based surrogate models is approximately done using that
of coarse models. This approach permits the creation of
library models that can be used for model enhancement
of microwave elements [5]. In some cases, the mapping established between parame-
ters of the coarse and fine models is not explicit and it is
hidden in the coarse model. This issue is addressed by
ISM. The drawback of this approach is that ISM technique
may not necessarily converge to the optimal solution.
This problem can be solved using the ISM along with the
response residual space mapping (RRSM) [6]. First, the
algorithm starts with the ISM to reach a solution close to
the optimal one. Then, RRSM approach is used to reach a
satisfactory solution.
In this paper, explicit SM modelling approach as well
as ISM and RRSM techniques are applied to a parallel-
coupled-line microstrip band-pass filter. Agilent ADS [7]
and Ansoft HFSS [8] are employed to simulate coarse and
fine models, respectively.
2
EXPLICIT SPACE MAPPING MODELLING The modelling procedure starts with optimization of the
coarse model to obtain the reference point of the region of
interest. According to star distribution, shown in Figure 1,
an
−
n
dimensional interval centered at the reference
point is created. Then, the input and output’s mapping
parameters are calibrated such that multiple sets of res-
ponses of the SM-based surrogate model match those of
the fine model, simultaneously. To check the validity of
the resulting model, it is tested with some test points in
the region of interest.
————————————————
• Tavakoli is with the Faculty of Electrical and Computer Eng., The Univer-
sity of Sistan and Baluchestan, Iran. • M. Zeinadini is with the Faculty of Electrical and Computer Eng., The
University of Sistan and Baluchestan, Iran • S. Mohanna is with the Faculty of