Reverse modeling of microwave circuits with bidirectional neural network models

Vai, Michael; Wu, Shuichi; Li, Bin; Prasad, S.

doi:10.1109/22.721152

Cited by 63 publications

(32 citation statements)

References 4 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…It was shown [2][3][4][5] that the process of searching for optimum parameters when designing a microwave circuit or device can be simplified when an artificial neural network (ANN) is used. Designing with ANNs makes use of the ANN propensity for storing knowledge, for generalizing patterns in data and for offering computational efficiency.…”

Section: Introductionmentioning

confidence: 99%

“…In conventional approaches [2][3][4][5] the ANN learns the relationships between input and output data at certain discrete frequency points. On the contrary, a major distinctiveness of the method presented in this Letter consists of training an ANN to provide a model, which describes a microstrip filter both in a continuous frequency band and in time domain.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Microstrip filter design using FDTD and neural networks

Banciu

Ambikairajah

Ramer

2002

Micro & Optical Tech Letters

View full text Add to dashboard Cite

better than 15 dB over the same range. The simulated phase difference between the two outputs, 88.2°Ϯ 1.7°, changes from 86.6°to 89.9°within the same band. With the same software the comparison between the conventional coupler and the proposed one shows 63% area saving, 300-MHz bandwidth reduction of isolation and return loss, and 2°maximum decrease in phase difference between the two coupling arms over the same range. EXPERIMENTAL RESULTSThe coupler designed at 900 MHz in the previous section was fabricated on Duroid dielectric substrate with r ϭ 2.2 and h ϭ 0.78 mm. The simple etching microstrip process has been used. The measured performances are shown in Figure 12. The measured S 21 is 3.7 Ϯ 0.4 dB between 840 and 950 MHz (110 MHz). Over the same range, the measured S 31 is 3.1 Ϯ 0.3 dB. The measure of return loss and isolation are better than 14 dB over the same range. CONCLUSIONA new method of reducing the size of 3-dB quadrature hybrid couplers has been presented. Design equations and curves have been introduced. In order to demonstrate the advantages of the design approach two quadrature couplers at two different frequencies, 0.9 and 10 GHz, have been designed and simulated. The area saving of the fabricated hybrid, at 900 MHz, has more than 65% relative to the conventional type. The proposed method does not need lumped elements or via-hole grounding, which leads to a low-cost, simple fabrication process and excellent design accuracy. These hybrids are suitable for MICs and MMICs applications. ACKNOWLEDGMENTThe authors wish to thank Professor Esmat A. F. Abdallah, Electronics Research Institute, Egypt for her useful discussion. Dr. Samir F. Mahmoud thanks Kuwait University for providing him with a sabbatical to do this work. MICROSTRIP FILTER DESIGN USING FDTD AND NEURAL NETWORKS

show abstract

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Microstrip filter design using FDTD and neural networks

Banciu

Ambikairajah

Ramer

2002

Micro & Optical Tech Letters

View full text Add to dashboard Cite

show abstract

“…In the optimization method, the EM simulator or the forward model is evaluated repeatedly in order to find the optimal solutions of the geometrical parameters that can lead to a good match between modeled and specified electrical parameters. This method of inverse modeling is also known as synthesis method [29]. Another method is to use an inverse model whose inputs are electrical parameters and outputs are geometrical parameters.…”

Section: Neural Network For Inverse Modeling Problemmentioning

confidence: 99%

“…The evaluation from input to output of a neural network model is also very fast. For these reasons, neural network techniques have been utilized in various microwave design applications [1]- [2], [4]- [5] such as vias and interconnects [6], embedded passives [7]- [8], coplanar wave-guide components [9]- [11], parasitic modeling [12], antenna applications [13]- [15], nonlinear microwave circuit optimization [16]- [18], nonlinear device modeling [19]- [22], power amplifier modeling [23]- [26], waveguide filter [27]- [29], enhanced EM computation [30], etc. In this article, we present an overview of neural networkbased modeling techniques and their applications in microwave modeling and design.…”

mentioning

confidence: 99%

Smart Modeling of Microwave Devices

Kabir

Zhang²,

et al. 2010

IEEE Microwave

205

118

View full text Add to dashboard Cite

Modeling and computer-aided design (CAD) techniques are essential for microwave design, especially with our drive towards first-pass design success. In the past few decades, tremendous progress in microwave CAD has led to a large variety of microwave models for passive and active devices and circuit components. The high quality and the availability of these models have enabled us to design circuits efficiently. These models have also allowed us to design larger and more complicated circuits than ever before.At the same time, new technologies and materials, emerging and non-traditional devices continue to evolve. Although the existing models are good for modeling mature technologies and existing devices, they are often inadequate or unsuitable when new devices are needed in system design. Conventional approaches to create or modify models are heavily based on slow trial-and-error processes.As new technologies and devices continue to evolve, we need not only new models, but also computeraided modeling algorithms such that model development becomes fast and systematic.At high frequencies, equivalent circuit models often lack fidelity. Detailed electromagnetic (EM) based simulations become essential to achieve design accuracy. However, EM simulations are computationally expensive especially when physical or geometrical parameters have to be repeatedly adjusted during design cycle. With the increasing design complexities, coupled with tighter component tolerances and shorter design cycles, there is a demand for design methodologies that are both accurate and fast at the same time. These are contradictory requirements and difficult to satisfy with conventional CAD techniques. The problem becomes even more severe in yield optimization and statistical validation where process variations and manufacturing tolerances of components are required to be taken into account. In addition, accurate parametric modeling techniques have become increasingly necessary, where we strive to describe not only the behavior of the microwave device, but also the change of the behavior against physical or geometrical parameters of the device.In recent years, neural network (NN) or artificial neural network (ANN) techniques have been recognized as a useful alternative to conventional approaches in microwave modeling [1]- [2]. Artificial neural networks can be used to develop new models or to enhance the accuracy of existing models. Neural networks learn device data through an automated training process, and the trained neural networks are then used as fast and accurate models for efficient high-level circuit and system design. These models have the ability to capture multi-dimensional arbitrary nonlinear relationships. The theoretical basis of neural network is based on the universal approximation theory [3], which states that a neural network with at least one hidden layer can approximate any nonlinear continuous multidimensional function to any desired accuracy. This makes neural networks a useful choice for device modeling where a mathemati...

show abstract

“…NNs are used in impedance matching [16,17], inverse modeling [18], measurements [19], and synthesis [20]. Multilayer perceptron (MLP), radial basis function (RBF), knowledge based neural network (KBNN), wavelet network, and recurrent neural network (RNN) are commonly used as ANN structures.…”

Section: Neural Network Developmentmentioning

confidence: 99%

Levenberg-Marquardt and Conjugate Gradient Neuro-Modeling of Simulated Miniature Rectangular Microstrip Antenna

Mishra¹,

Janvale²,

Kasar³

et al. 2012

Advanced Computational Techniques in Electromagnetics

View full text Add to dashboard Cite

Neural-network computational modules have recently gained recognition as an unconventional and useful tool for RF and microwave modeling and design. Neural networks can be trained to learn the behaviour of passive/active components/circuits. This work presents artificial neural network (ANN) for design of IE3D simulated miniature microstrip antenna. In the presented work, the artificial neural network is used for accurate determination of different parameters like resonant frequency, bandwidth, return loss, and voltage standing wave ratio (VSWR) of square and rectangular microstrip patch antenna. The developed neural network model which uses the data of simulated hundred antennas is based on Levenberg-Marquardt (LM) and conjugate gradient (CG) feed-back propagation. The developed ANN models for rectangular microstrip antennas (RMSAs) are in very good agreement with the experimental results available in the literature. The comparative analysis of developed models is presented which gives higher accuracy than that reported elsewhere.

show abstract

Reverse modeling of microwave circuits with bidirectional neural network models

Cited by 63 publications

References 4 publications

Microstrip filter design using FDTD and neural networks

Microstrip filter design using FDTD and neural networks

Smart Modeling of Microwave Devices

Levenberg-Marquardt and Conjugate Gradient Neuro-Modeling of Simulated Miniature Rectangular Microstrip Antenna

Contact Info

Product

Resources

About