The paper proposes a classification of sound sources in virtual environment systems, methods for creating such sources in the 3DS Max three-dimensional modeling system, and methods for controlling them using virtual control panels and their corresponding functional circuits. 

The development of a pulse step-down voltage converter begins with the calculation of the parameters of the inductor and the output capacitor. The value of the inductance of the inductor and the capacitance of the output capacitor depend on many characteristics of the converter, such as the power of the converter, the values of the input and output voltage, the permissible level of output voltage ripple, the permissible relative pulse of the output current, and the modulation frequency. As a rule, the characteristics of the converted voltage and the power of the converter are set by the technical specification and cannot be independently selected by the developer. The modulation frequency is not one of the parameters characterizing the output power of the converter and can be selected by the developer at his discretion. The values of the inductance of the inductor and the capacitance of the output capacitor have a significant dependence on the selected modulation frequency. The higher the modulation frequency, the lower the inductance and capacitance of the output capacitor of the converter. This minimizes the overall physical dimensions of the pulse converter. However, as the modulation frequency increases, energy losses on the power key increase and the effect of electromagnetic interference increases. Choosing the modulation frequency of a pulse voltage converter is always a search for a compromise between minimizing the size of the inductor with the output capacitor and reducing the efficiency of the converter, along with reducing the noise immunity of the powered device. This article provides an example of the analysis of the main parameters of the converter's power circuit in a wide range of modulation frequencies. These calculations help the developer to find the optimal ratio of the modulation frequency with the values of the inductor and the output capacitor, if different modulation frequencies are allowed in accordance with the technical specification. An example of calculating the dependence of the parameters of a pulse converter on the output voltage and output power at a fixed modulation frequency is also presented.

A CMOS 12-bit analog-to-digital converter of conveyor type with design standards of 28 nm is presented. Comparatively low power consumption with a sufficiently high sampling frequency of up to 125 MHz is achieved through the use of fully differential two-stage operational amplifiers in the pipeline architecture, operating in a 1.5-bit ADC pipeline architecture per stage with a supply voltage of 1.8V.

A thermal model for silicon field-effect GAA nanotransistors is discussed, taking into account the thermal effects caused by boundary roughness. The model is based on the perturbation theory method, which takes into account the effect of the dependence of the nanowire diameter and surface roughness on the thermal conductivity of the transistor channel, as well as the effect of the characteristics of the GAA nanotransistor structure on heat dissipation. In this case, for thermal correction of the (idealized) transistor model, for which a proven mathematical model has been developed, a weak perturbation should be added through an additional "perturbing" Hamiltonian. Then the various physical quantities associated with the perturbed system can be expressed as "corrections" to the characteristics of the original model. These corrections are a priori small compared to the size of the quantities themselves. However, they 1) significantly change the characteristics of the initial system, and 2) simplify the algorithms for their calculation. Based on the model, the influence of the nanotransistor chip design parameters on its heat dissipation is discussed. The developed model can be used to design circuits based on nanotransistor chips, taking into account thermal factors.

The article is devoted to the use of OpenCL technology, which allows using the powerful resources of graphic processors to improve the performance of computing programs. Techniques for developing effective parallel programs in the OpenCL environment are considered to speed up the convolution operation for multiple complex vectors based on the fast Fourier transform.

The task of planning computing in multiprocessor systems is considered in periodically incoming requests to perform a set of work with non -fixed durations. A heterogeneous set of resources is used - renewable and non -renewable. There are cases when work allow interruptions and switching from one processor to another, as well as when the work is uninterrupted. Algorithms have been developed in which, when processing each request, the schedule built during the processing of the previous request is adjusted. Algorithms are based on network modeling and searching for the maximum flow and flow of the minimum cost.

The task of detecting plagiarism in programming task solutions holds high priority in digital educational platforms due to the necessity of providing accurate and reliable assessment of users' learning progress. Methods that compare submitted solutions solely based on textual similarity, without accounting for the syntactic characteristics of the programming languages in which the solutions are written, often fail to deliver precise and trustworthy results—similarly to statistical approaches based on machine learning. This study proposes a method for block-based comparison of student programming solutions for plagiarism detection, taking into account the syntactic features of programming languages.

The paper addresses the task of automating student attendance tracking within the Mirera digital educational platform. Two innovative approaches have been developed and investigated: time-based presence confirmation using QR codes with Server-Sent Events and JWT tokens, and the introduction of a control metric through student video identification based on 3D classroom reconstruction. Experimental results demonstrate that the proposed solutions enhance the accuracy and efficiency of attendance tracking, providing reliable control over the educational process in a hybrid learning environment.

We constructed and studied a model of interaction between a community of relatively weak prey  agents and a strong predator agent in a two‑dimensional grid world (a lattice environment typical of grid automata and agent-based models). The predator can attack, kill, and consume prey agents. Each prey agent is controlled by a neural network and adopts one of two behavioral strategies: (1) normal activity, or (2) defense against the predator. In the normal activity strategy, prey agents lie dormant, feed, breed, and move through the grid. In the defense strategy, they attempt to escape, threaten, or attack the predator. The neural network outputs control each agent’s ac tions. The predator follows a simpler, rule-based protocol: it can lie dormant, evade threatening prey, or attack them. Its behavior is governed by basic logic. We analyzed the model using computer simulations. We found that, with realistic parameters, the prey agents collectively overcome the predator: prey resource levels increase steadily, while the predator’s resources decline to zero, leading to its extinction. We also discovered that successful defense requires a sufficiently abundant food supply; when prey food is scarce, the predator successfully suppresses the prey population. We used computer simulation to analyze the model. When the prey agents’ food supply is low, the predator agent suppresses the prey agents.