Monday May 13, 2019, 8:30-10:30
Prague Congress Centre, North Hall

Workshop: "Advances in Smart Modeling Techniques for Microwave Engineering"
Workshop Organizers:
Zlatica Marinković, University of Niš, Serbia
Giovanni Crupi, University of Messina, Italy

In the past two decades, smart techniques based on the artificial intelligence have found their place as a very efficient and powerful tool for modeling and optimization in the field of microwave engineering. Having in mind a growing general interest for the artificial intelligence, the aim of this workshop is to give an insight into the state-of-the art and future directions in the field of smart modeling techniques for different microwave applications. The majority of presentations will refer to the applications of artificial neural networks, as the most frequently used smart technique in the field of microwaves, for modeling and optimization of passive and active microwave devices, as well as other applications such as spatial localization of microwave sources. Moreover, the workshop will cover applications of other smart techniques, among them support vector machine methods, symbolic regression, and evolutionary algorithms.

8:30-9:10: "Artificial Neural Networks for Microwave Design-Recent Advances and Future Directions" 
Q. J. Zhang
Carleton University, Canada

Artificial Neural Networks (ANNs) are information processing systems inspired by the ability of human brain to learn from observations and to generalize by abstraction. ANN has been recognized in the microwave CAD community as a powerful tool to address the modeling and computational challenges in linear and nonlinear microwave modeling and design. This talk presents an overview of this area and describes recent advances in new training methods, knowledge-based modeling, automated model generation, model extrapolation, inverse modeling and multiphysics oriented parametric modeling. This talk also provides a discussion on the trends and future directions of ANN for microwave design.

9:10-9:30: "Artificial Neural Networks for Small-Signal Modeling of Advanced Microwave Transistors: Does the Device Technology Influence Modeling"
Zlatica Marinković*, Giovanni Crupi**, Alina Caddemi**, Vera Marković*, Dominique M. M.-P. Schreurs***,
* University of Niš, Serbia, **University of Messina, Italy, ***KU Leuven, Belgium

Artificial networks (ANNs) have been widely used for modelling the small signal scattering parameters of microwave transistors. This presentation is meant to give a comprehensive overview on the high-frequency small-signal transistor modelling by considering various cases of study. Different device technologies will be analyzed (GaAs HEMTs, GaN HEMTs, Si FinFETs, Si FinFET varactors, and Si MOSFETs) under many operating conditions: frequency, bias point, ambient temperature, and geometrical dimensions. Based on this comparative analysis, the presentation will focus on examining how the device technology and the operating conditions influence the ANN model construction and performance.

9:30-9:50: "New Applications of Artificial Intelligence in Microwaves"
Filiz Güneş*, Mehmet Ali Belen**
*Yildiz Technical University, Turkey, **Artvin Coruh University, Turkey

Herein, new major applications of Artificial Intelligence (AI) in microwave theory and techniques in the last decade are presented. First of all, "Performance Data Base" of a microwave transistor can be built using learning machines and intelligent optimization algorithms without any need of expert knowledge. This can be considered as a reform. Thus this facilitates design of LNA with the full potential of the active device with the minimum consumption and optimum performance. Furthermore, design of small-signal microwave amplifier will be possible completely with AI using knowledge based SV modelling microwave transmission line in addition to the Performance Data Base. Big data analysis and analytical formulation is also taken place. Design and optimization of antennas is also taken place as an import advance as AI application.

9:50-10:10: "Forward and Inverse Neural Modeling of RF MEMS switches"
Tomislav Ćirić*, Rohan Dhuri**, Olivera Pronić-Rančić*, Vera Marković*, Zlatica Marinković*,
*University of Niš, Serbia, **ALTEN GmbH, Munich, Germany

The increased growth of the applications of RF MEMS switches in modern communication systems has created an increased need for their accurate and efficient models. Artificial neural networks have appeared as a fast and efficient modeling tool providing similar accuracy as the standard commercial simulation packages. This talk gives an overview of the applications of artificial neural networks in modeling of RF MEMS capacitive shunt switches. The most important switch characteristics in electrical and mechanical domains versus the bridge dimensions are modeled. Moreover, inverse models aimed to determine the switch bridge dimensions for given requirements for the switch characteristics are developed. Also, an inverse modeling approach based on combination of artificial neural networks and a lumped element circuit model is considered.

10:10-10:30: "Neural Network-based Approaches for the Spatial Localization of Microwave Sources with Stochastic Radiation"
Zoran Stanković, Nebojša Dončov, Biljana Stosić
University of Niš, Serbia

A brief overview of the results achieved in the field of application of artificial neural networks for a spatial localization of stochastic electromagnetic (EM) radiation sources is given. The theory of stochastic sources' near-field representation in the form of antenna arrays with dipole elements and correlation matrix calculation in the far-field scan area as well as the architecture of developed neural models for the 1D and 2D DoA (direction of arrival) estimation are briefly presented here. Different scenarios are considered: 1D scenario - the stochastic sources are moved in the azimuthal plane and their position is determined by only one incidental angle in relation to the linear antenna array, and 2D scenario - the stochastic sources are moved in a plane that is parallel to the rectangular antenna array so that their position is determined with two spatial angles. In addition, a few examples are given to illustrate the presented neural models' capabilities to efficiently perform the DoA estimation for either uncorrelated or partially correlated sources.

Presenter Contacts:
Qijun Zhang:
Zlatica Marinković:
Filiz Güneş:
Tomislav Ćirić:
Nebojša Dončov:

Monday May 13, 2019, 12:30-15:30
Prague Congress Centre, North Hall

Workshop: "Active, Non-Foster, and Time-varying Electromagnetic Structures and System"
Workshop Organizers:
Silvio Hrabar, University of Zagreb, Croatia
Dmitry Kholodnyak, Saint Petersburg Electrotechnical University "LETI", Russia

Usually, the electromagnetic and electronic parts of standard communication system, such as antennas, amplifiers, oscillators etc. are analyzed and designed separately. However, the recent progress of active electromagnetic structures and systems is changing this approach and merging the fields of electromagnetics and RF/microwave engineering. Here, the electronic and electromagnetic parts are joined into an entity with novel physical phenomena such as negative, switching, or time-varying capacitance/inductance. Applications associated with this novel approach include advanced wireless power transfer systems, broadband/ non-reciprocal/ self-oscillating antennas, metasurfaces, filters, hybrids and phase shifters, just to mention a few. This workshop gathers together a team of prominent experts that will give an overview of both the background theory and applications of this exciting research field.

12:30-12:35: Introduction

12:35-13:05: "Wireless Power Communications Based on Backscatter, an Alternative for IoT Sensor Development"
Nuno Borges Carvalho, Universidade de Aveiro, Portugal

13:05-13:35: "Non-reciprocity and Control of Doppler Effect in Antenna Systems Induced by Active Time-varying Metamaterials and Metasurfaces"
Davide Ramaccia, Alessandro Toscano, and Filiberto Bilotti, Università Degli Studi Roma Tre, Italy

13:35-13:50: Break

13:50-14:20: "Applications of Non-Foster Elements in RF/Microwave Engineering – Basic Ideas and Background Physics"
Silvio Hrabar, University of Zagreb, Croatia

14:20-14:50: "Non-Foster Forms and Their Application to Small Antenna Designs"
Daniel Segovia-Vargas and Fernando Albarracín-Vargas, University Carlos III de Madrid, Spain

14:50-15:20: "RF/Microwave Devices with Non-Foster Elements"
Dmitry Kholodnyak, Saint Petersburg Electrotechnical University "LETI", Russia

15:20-15:30: Conclusion and Discussion

Presenter Contacts:
Nuno Borges Carvalho:
Davide Ramaccia:
Silvio Hrabar:
Daniel Segovia-Vargas:
Dmitry Kholodnyak:

All material Copyright ©2019 Horizon House. All rights reserved.