Оценка искажений диаграммы направленности радиолокационной станции при нестационарном воздействии тепловых процессов на малошумящий усилитель приемного тракта / А. В. Тимошенко, А. С. Захаров, И. А. Байкалов [и др.]. - Текст : непосредственный : электронный
// ЦИФРОВАЯ ОБРАБОТКА СИГНАЛОВ. - Москва : РНТОРЭС им. А. С. Попова, 2025. - № 1. - С. 22-26. - URL: https://elibrary.ru/item.asp?id=80597078 (дата обращения: 29.04.2025). - URL: http://www.dspa.ru/abstracts/ab-125.php#an125_4 (дата обращения: 29.04.2025).

Аннотация: Modern radar stations strive to increase the resolution to accurately determine the coordinates and characteristics of objects. One of the promising methods to achieve this goal is the use of a super-resolution mode. However, thermal distortions occur during the operation of the radar, which significantly reduce the efficiency of superresolution. This article presents a new model for the analysis and compensation of these distortions, based on the use of the original integro-differential equation. It is shown that the unsteady behavior of thermal processes leads to distortions of the amplitude-phase distribution, amplitude-frequency response and radiation pattern, which changes the resolution characteristics. It is proposed to compensate for the influence of thermal processes by means of operational calibration of amplitude and phase, taking into account the distortions of the radiation pattern. A demonstration experiment was conducted to simulate the radiation pattern at different temperatures of low-noise amplifiers of the receiving path and the same heating rate

Алимагадов, К. А. (Автор МИЭТ (аспирант); Alimagadov, K. A.).
Нейросетевая винеровская фильтрация изображений в области дискретного вейвлет-преобразования / К. А. Алимагадов, С. В. Умняшкин. - Текст : непосредственный : электронный
// ЦИФРОВАЯ ОБРАБОТКА СИГНАЛОВ. - Москва : РНТОРЭС им. А. С. Попова, 2025. - № 1. - С. 39-46. - URL: https://elibrary.ru/item.asp?id=80597081 (дата обращения: 29.04.2025). - URL: http://www.dspa.ru/abstracts/ab-125.php#an125_7 (дата обращения: 29.04.2025).

Аннотация: The research is devoted to quality improvement of additive white Gaussian noise suppression in images. The architecture of neural network for image denoising based on Wiener filtering in the domain of discrete wavelet transform (DWT) has been developed. The key point of the processing scheme is the adaptation of image local filtering technique based on contour analysis in the domain of DWT. The number of trainable parameters of the proposed network is an order of magnitude smaller as compared with the modern machine learning models used to solve the same task. The neural network was tested on «Berkeley segmentation data set 500». Estimation of filtering quality was based on analysis of peak signal to noise ratio and structure similarity index values. It was found that the developed approach provides higher quality of noise suppression than other considered algorithms based on Wiener filtering