Ilya Pershin: “Miniature antennas are required for various applications”
4 October 2021
Ilya Pershin, a second year master student of the Institute of Computational Mathematics and IT, is among the winners of UMNIK Sirius 2021 grant competition with a prize of 500 thousand rubles. His project concerns modeling of complex geometry antennas. As Pershin says, he has been working on the topic since his third year, “I turned to my supervisor, Dmitry Tumanov, who advised me to start researching this topic. The task was to simulate the electrodynamic characteristics for a given antenna geometry. Electrodynamic characteristics allow you to understand in which direction the antenna will most effectively receive and radiate signals; at what resonant frequency will it work; what is its bandwidth. In practice, we are interested in the inverse problem – to obtain the corresponding geometry from a set of electrodynamic characteristics. The complication of the antenna geometry makes it possible to improve its electrodynamic characteristics, approach the target parameters with a smaller error and reduce its dimensions.”
His project is now set to take two years, “In the first year of work, the most difficult part is planned – the solution of the inverse problem. The second year of work is aimed at the implementation of these tasks in the form of a boxed solution. There will be two options for implementing the project: the first is the classic one, when a client buys a license and installs software on their computers. This approach has a number of disadvantages. The most significant is the lack of computing power of the client, so the second option was proposed – the development of a web platform. In this option, calculations will take place on project resources.”
The student says that such an approach may find its way to the smaller businesses, “Miniature antennas are required in various applications. For example. phones and laptops. It is possible to reduce the antenna and preserve its electrodynamic characteristics by complicating the geometry. I suggest using a combination of classical antenna modeling techniques and machine learning techniques. The advantage of the web platform lies in the possibility of using a flexible tariff plan. We can provide the client not with the purchase of a whole license for a year but with the introduction of some tariff plans – if they need to calculate five antennas, they will pay for five. This can be very beneficial.”