Thesis supervisor: Sándor Bilicz
Location of studies (in Hungarian): Department of Broadband Infocommunications and Electromagnetic Theory Abbreviation of location of studies: SZHVT
Description of the research topic:
The simulation of propagation and scattering of electromagnetic waves is essential in the analysis, design and optimization of a wide class of electrical devices. More precisely, in this thesis we focus on the case of electrically large scatterers and propagation environments, in which the asymptotic approximations (e.g., physical optics, ray tracing, homogenization) of wave phenomena are needed.
These simulations are reliable when simple scenarios (e.g., canonical-shaped scatterers, homogeneous media, far field conditions, etc.) are considered. However, the computational complexity may be large even in such cases. The extension of the scope of the simulations to complex configurations, hence to better model real-life applications, is even more challenging.
Algorithms included in the wide class of machine learning methods can potentially be useful in the solution of the following problems.
(i) Building a surrogate model that operates at a low computational cost and provides a good approximation of time-consuming electromagnetic simulations.
(ii) Predicting the output of a scattering problem in a complex scenario, based on the knowledge of simulated outputs in elementary scenarios, to which the complex scenario can be traced back. This may include the decomposition to spatial sub-problems or the consideration of time-varying environments based on static simulations.
(iii) Better understanding how the different features of a scenario are reflected in the output data. What are the most influential features, and how does the uncertainty of the configuration influence the output uncertainty.
The main aim of this thesis is to cast the above problems as regression, interpolation and pattern recognition tasks, and combine electromagnetic simulations with machine learning methods in a proper way. As potential applications of the expected results, one may mention millimeter-wave radars in driving-assistance systems, or propagation of 5G telecommunication signals.
Required language skills: English Number of students who can be accepted: 1
Deadline for application: 2021-09-01
2024. IV. 17. ODT ülés Az ODT következő ülésére 2024. június 14-én, pénteken 10.00 órakor kerül sor a Semmelweis Egyetem Szenátusi termében (Bp. Üllői út 26. I. emelet).
Login
