The influence of pulsed photon processing in a nitrogen medium with an incoherent radiation flux from quartz halogen lamps directed to the non-working side of the wafer, providing heating to 1150 °C in about 7 s, on the optical properties of 17.7 nm thick silicon dioxide layers formed by pyrogenic oxidation of silicon doped with boron with the orientation (100), and on the electrophysical characteristics of the interface with silicon was established using the methods of infrared Fourier spectrometry, spectral ellipsometry, time-of-flight mass spectroscopy of secondary ions, and studies of the current-voltage and capacitance-voltage characteristics. It was revealed that pulsed photon processing of silicon dioxide layers leads to compaction and rearrangement of its structure, as well as to the formation of Si‒N bonds in silicon dioxide, providing nitridation of SiO2. This is indicated by the shift, decrease in the half-width and voltage of the main absorption band of the Si‒O bond, decrease in the refractive index from 1.48 to 1.47 and increase in the layer thickness to 18.2 nm. It is shown that nitridation of SiO2 layers during pulsed photon processing in a nitrogen atmosphere leads to a decrease in the leakage current of the dielectric by four times and its charge density by 3.43 times. This is 2.19 and 3.01 times more, respectively, than during processing in natural atmospheric conditions due to the formation of a layer with an increased nitrogen concentration at the Si–SiO2 boundary. The results obtained can be used to create dielectric coatings in electronic products.

Aluminum oxide and barium titanate films were formed on titanium and oxidized single-crystal silicon substrates using the sol-gel method. The dependence of the capacitance and the dielectric loss tangent on the frequency of the Al2O3/BaTiO3/Ni capacitor structure was studied in the range of 200 Hz–200 kHz on a titanium substrate using the substrate as the lower electrode. In the high-frequency region (20–200 kHz), the standard deviation of the capacitance was 95–97 pF, and in the low-frequency region (0.2–10.0 kHz), it was 80–94 pF. The capacitance and dielectric loss tangent values remained within the established standard deviation values for a year. Examples of images of microdisks and waveguides obtained by photolithography and chemical etching of structures with Al2O3, SiO2 and BaTiO3 films are given for subsequent development of planar waveguides and electrooptical modulators.

The optical processes in the structure of a thin-film IR LED have been simulated using the finite difference method in the time domain. The parameters such as transmission and propagation efficiency of electromagnetic waves in the range of 1.2–1.4 μm have been investigated. A control device with an indium tin oxide (ITO) layer as a transparent conducting electrode has been simulated. The replacement of the ITO layer with a fluorine-doped tin oxide (FTO) layer has been considered. It has been found that with such a replacement, the transmittance of IR radiation passing through the FTO functional layers increases to 70 %, and the angular distribution of E2 increases by 10 compared to a device with an ITO layer. Thus, it is advisable to replace the layer of the transparent conducting ITO electrode with the FTO layer.

The relevance of the study is due to the rapid development of microelectromechanical systems (MEMS), which are used as logic elements due to low power consumption, high reliability and the ability to reprogram during operation. However, existing approaches to the design of MEMS logic gates require the development of compact and accurate mathematical models to analyze their dynamic characteristics and logical operations at the system level. The article presents a method for synthesizing parametric mathematical models of MEMS logic gates based on silicon resonators with electrostatic excitation. A library of design macromodels has been developed, including three-, fiveand nine-pin MEMS resonators, which allow analyzing static, frequency and dynamic characteristics. The analysis of the electrode collapse effect, harmonic analysis and simulation of logical operations such as NOT, AND, OR-NOT and exclusive OR are carried out. The developed models can be used for designing analog-todigital and digital-to-analog converters, triggers and arithmetic logic devices based on MEMS technologies.

The paper presents a technique for analyzing the characteristics of the electromagnetic background near the earth’s surface generated by emissions from megaconstellations of low-orbit communication satellites (LOCSS) based on estimates of the average territorial density of wireless traffic generated by them on the earth’s surface and available system characteristics of the LOCSS and their constellations, such as orbit altitude, characteristics of antenna patterns, limitations on the angle at which the main lobe of the LOCSS antenna can be directed to the earth’s surface, and features of subscriber terminal servicing scenarios. The obtained estimates of the average levels of the electromagnetic background, corresponding to the range of real values of the average density of wireless traffic generated by LOCSS constellations on the earth’s surface, significantly exceed the levels of the natural electromagnetic background. This is consistent with the estimates obtained earlier using data on the total radiated power of LOCSS and their number in the megaconstellation, and confirms the adequacy of the presented technique.

The possibility of using large language models, generative machine learning and big data methods for predictive analysis of the electronic properties of nanostructures based on semiconductor materials is considered, without limiting the generality of this approach for other crystalline materials. The conceptual solution in the form of a cross-platform software application with an advanced search generation capability for applying this approach to working with scientific data arrays is described. Preliminary test results showed a positive result of using the developed solution to predict semiconductor properties (bandgap width, Fermi energy) of a reference material. Prospects for the development and implementation of big data methods, large language models, and generative artificial intelligence in the context of modern trends in materials science are discussed.

The technique of automatic generation of questions and answers for knowledge testing is considered. The proposed approach is based on the developed method of defining and using clusters of keywords to generate questions and answers based on a global or local language model, which constitutes theoretical and applied novelty. Clusters are built on the basis of a corpus of text documents related to the studied area (textbooks, methodological manuals, electronic resources). The technical solution presented in the article is logically complete and can serve as a basis for practical developments.

The concepts of QoE (quality of experience), QoS (quality of service), and GoS (goal of service) are commonly discussed in network performance and user satisfaction studies. Although QoE has become a popular topic in research, the boundaries between QoS and QoE are often blurred, making it difficult to clearly define them. This paper examines the differences and relationships between these terms with regard to their practical applications. QoS is a subjective metric that reflects how users perceive a service. It is influenced by personal preferences and various environmental factors. GoS measures the probability of a successful connection or call under certain conditions. The results showed that implementing QoS features such as traffic prioritization can positively affect both GoS and QoE by reducing packet loss and improving service reliability. It is shown how network performance management using QoS tools can improve user experience and overall service reliability, providing a clearer understanding of how these concepts interact in practice. 

The article considers an approach to determining the similarity of datasets for training algorithms using datasets with human faces as an example. This approach allows finding similar datasets from different sources, expanding the detection of features and classes and significantly affecting dataset balance. For each dataset object, a vector representation (embedding) was obtained, then the embeddings in both datasets were compared. The experiments were conducted using datasets with images of human faces as an example. To obtain embeddings, a pretrained ResNet network was used. During the research, one dataset was divided into two parts, which were similar datasets, then each of the parts was compared with a different dataset. The new similarity metric is proposed, which has several advantages and allows to find the most similar datasets.

The article presents an assessment and analysis of uncertainties in pathological lesion visualization on PET/CT images that arise due to respiratory movements of a biological object, which have a direct impact on determining the geometric characteristics of the pathological lesion, its localization, and the correctness of modeling the three-dimensional distribution of the radiation dose in the patient’s body. Based on the experimentally established dependencies of the influence of the displacement value and the diameter of the object (sphere) under study on the value of the discrepancy between the visualized volume of the object under study, a mathematical model of respiratory distortions of the pathological lesion during visualization on PET/CT images has been developed, which allows for a quantitative and qualitative assessment of the influence of the patient’s respiratory movement on the geometric accuracy of pathological lesion visualization on PET/CT images with up to 98 % accuracy. Verification of the model has shown a high degree of consistency between the model calculations and experimental data. The average values of relative uncertainty for CT and PET, respectively, were (2.123  1.051) % and (2.661  0.870) %, and for absolute uncertainty – (2.096  0.941) % and (1.992  0.782) %, which confirms the correctness and practical applicability of the developed method.

The scientific novelty of the article is the technology of using a medical blockchain based on a smart contract to support patient therapy after IT diagnosis of Alzheimer’s disease. In combination with the subsystem of IT diagnostics of a patient, the exchange of recognition data for the degree of Alzheimer’s disease is implemented. The smart contract performs automated management of therapy support for patients with diagnosed Alzheimer’s disease . The functional structure of the therapy support subsystem (smart contract in the Etherium blockchain) has been developed, including initialization of patient data, treatment plans, registration of medical staff, record management, dynamic switching of therapy plans based on diagnostic data, monitoring of medical staff warnings, as well as management of payments and their distribution. The proposed smart contract enables a new approach to interaction and collaboration in the field of therapy management and can be used for neurological patients with other diseases.

The main goal of radiation therapy is to deliver a dose of ionizing radiation to a tumor. When irradiating foci located on the skin surface or at a shallow depth, the problem of overirradiation of healthy tissues arises. In order to optimize the distribution of the dose of ionizing radiation, auxiliary devices in the form of boluses are used in treatment practice. In this case, correct fixation of the bolus, ensuring its tight fit, is of great importance. An increase in the quality of dose distribution with tight fixation of the bolus has been experimentally proven. Numerical values of the dose at the control point with different bolus fit density have been obtained using different methods for calculating irradiation plans.

A complex encryption algorithm has been developed based on traditional RSA (public key encryption algorithm), SHA-3 (hashing algorithm), and a discrete quantum map. The features of the proposed algorithm and the developed computer program in C++ are described. To evaluate the performance of the algorithm, testing was carried out on three encrypted images that differed in the volume of transmitted information. A comparative analysis of the obtained images with the original ones showed the efficiency of the information encryption algorithm using a discrete quantum map. This algorithm can also be used to perform additional operations related to including information inside the encrypted image, which will provide an additional level of protection when transmitting information.