Bulletin of Voronezh State Technical University

this article addresses the current problem of ensuring data security during transmission and storage. The proposed algorithm is designed to operate in a stream mode using cascade (double) encryption, which enables efficient encryption of data streams without creating and storing a file with the original data on the disk; it also eliminates the need for re-reading the information being encrypted. The algorithm is based on the use of two stream ciphers (ChaCha20 and HC-256) and key derivation functions - Argon2 and scrypt. The key features of the algorithm are: the use of cascade encryption with stream ciphers combined with key derivation functions to ensure the uniqueness of encrypted data even when the original data is identical; the key encryption mechanism allows for changing passwords without the need for complete data re-encryption. The object of consideration in this article is an algorithm that provides a high level of protection, where an attacker would need to simultaneously compromise two independent keys to access the data, which is computationally prohibitive even with modern equipment. The proposed solution is intended for the secure long-term storage and transmission of confidential information, such as passwords, personal, and other sensitive data

under the assumption of ideal flow in a single-pass cross-flow plate heat exchanger of a "hot" coolant with volumetric heat release and a "cold" coolant with constant thermophysical properties and no thermal resistance of the heat-transfer surface, analytical expressions for the temperature fields of coolant flows in the heat exchanger are obtained using a one-sided integral Laplace transform. The system of partial differential equations with initial conditions is classified as a 2-D Cauchy problem on a semi-bounded quadrant, and the solution is presented in quadratures using special Bessel and Dirac functions. In the absence of heat release, the solution coincides with the Schumann solution. A computational experiment has shown that failing to account for the heat release of the "hot" coolant can lead to a significant error in determining the temperature at the heat exchanger outlet and, ultimately, to an incorrect value of the heat transfer area. Based on the analytical solution obtained for the heat transfer problem in a single-pass cross-flow plate heat exchanger, under the assumption of a hydrodynamic regime close to the ideal plug flow regime (in a first approximation, corresponding to turbulent flow), a method for determining the main thermohydraulic characteristics that ensure a given cooling regime for the heat-generating medium has been developed. The presented method is demonstrated using a specific example demonstrating the significant influence of the volumetric heat release in the cooled coolant on the geometric characteristics of the cross-flow plate heat exchanger and, accordingly, on the heat transfer area, which can ultimately lead to a decrease in the heat exchanger's efficiency

movement planning is important for the smooth performance of the tasks of autonomous mobile objects (AMO), and is also the subject of close attention by scientists from all over the world. The AMO chooses the optimal or suboptimal route from the start to the end point in order to meet pre-defined target requirements while ensuring its own mobility and in the face of external interference. Despite the fact that many studies have improved the trajectory planning method, its convergence is still an urgent topic, since the complexity of the algorithm is high and it is easy to get into local optimum. Within the framework of this work, a new approach to solving this problem is proposed – a multi-criteria optimization algorithm based on navigation variables (MCNV). For the first time, the route is planned as a multi-criteria optimization task using a set of objective functions reflecting the conditions of optimality and safety of AMO management. MCNV is used to minimize these functions by searching for Pareto optimality. The algorithm implements a new trajectory representation based on navigation variables, which allows taking into account the kinematic limitations and dynamic capabilities of the AMO. In addition, an adaptive transformation mechanism is used to increase the diversity of solutions and obtain more effective trajectories. The simulation results showed that the MCNV method has the ability to generate a variety of non-dominant solutions, which makes this algorithm a universal tool capable of meeting various requirements of applied tasks, and increases the likelihood of successful planning

An approach to interpreting experimental data on the dynamic viscosity of water-based nanofluids taking into account the polydispersity of nanoparticles is presented. Existing theoretical concepts of the dynamic viscosity of nanofluids do not provide complete information on its dependence on the degree of polydispersity of nanoparticles. This is due to the fact that single-stage and two-stage methods of nanoparticle generation cannot currently be controlled using instrumentation due to their small characteristic scale. In addition, there are such complicating factors as Brownian diffusion, migration flow, thermophoresis, diffusiophoresis, and sedimentation. To date, experimental data on the demonic viscosity of liquid nanosystems do not reflect the granulometric composition of the nano-part, but instead indicate the median component. It is proposed to describe the degree of polydispersity of nanoparticles by the classical distribution function of the dependence of concentration on the size of dispersed phase elements and to identify a conjugate relationship with a similar counting function of their distribution. This made it possible to obtain a modified expression for predicting the dynamic viscosity of nanofluids based on an empirical relationship generalizing a large array of experimental data, followed by special normalization for the monodisperse case and averaging over a size range. Assuming that the distribution of nanoparticles by size is in accordance with Gaussian laws, it is shown that with increasing dispersion, the dynamic viscosity of nanofluids decreases and, therefore, it is possible to reduce energy losses for transportation in various technical and technological systems by controlling the granulometric composition of nanoparticles.

The influence of the dimensions of geometric elements and the spatial orientation (tilt angle to the build platform) of cylindrical products on the accuracy of additive printing using the FDM (Fused Deposition Modeling) method is considered. The relevance of the study is due to the growing use of 3D printing in various industries, including small-scale production, where the shape of products is constantly changing. Quantitative values ​​of absolute and relative deviations of sizes from a given shape were experimentally obtained for cylindrical products with a diameter range from 3 to 13 mm with a step of 2 mm and heights from 5 to 30 mm with a step of 5 mm at tilt angles from 0° to 90° with a step of 15°, using PETG (polyethylene glycol) material on an Elegoo Neptune 4 3D printer. To obtain intermediate values ​​of deviations and shape distortions within the range of measured experimental data, a mathematical model based on sequential monotonic cubic interpolation using the Fritch-Carlson algorithm was developed. The obtained results can be used to optimize printing accuracy and improve product quality in the field of 3D printing.

The problem of constructing an engineering model of battery operation suitable for its practical use in studying control processes in autonomous power supply systems is solved. A battery with promising sealed nickel-hydrogen batteries that do not require direct maintenance during operation is selected as a specific object of study. It is shown that the heat-and-power part of the general model of such a battery operation can be represented by a first-order inhomogeneous differential equation with a nonlinear right-hand side determined by the internal heat release of the batteries during charging and discharging. To approximate the noted nonlinearity, a quadratic function is proposed, its parameters are identified based on experimental data, and a numerical solution of the differential equation is given. For the electric power part of the model, the basic Nernst algebraic model is used, which includes terms reflecting Ohmic voltage losses in the battery, as well as a power-law factor determining the nonlinear effect of the battery charge state on its voltage. The relationship between the thermal and electrical parts of the model is implemented by including current in the thermal part of the model, and temperature in the electrical part. The constructed model was used to study the processes of control the storage battery in the nominal mode of alternate charge and discharge, in the cyclic mode with insufficient charge and in the mode of long-term charge storage in the cooled state of the battery. The results of these studies are presented, confirming the adequacy of the proposed models.

This paper explores methods for optimizing programmable gate array (FPGA) architectures for the efficient implementation of neural network decoders for block codes, including low-density parity-check (LDPC), polar, and Bose-Chaudhuri-Hocquenghem (BCH) codes. The focus is on hardware modifications that enable achieving an optimal balance between decoding accuracy and computational efficiency. Specialized architectural solutions for FPGAs are developed and analyzed, including modified lookup table (LUT) schemes with adaptive bit depth and hardware accelerators for real-time node checking operations. The developed solutions are particularly relevant for iterative decoding algorithms (Min-Sum) applied to LDPC codes, which require intensive soft-decision processing. Experimental results demonstrate a 30% increase in decoder throughput compared to traditional implementations, a 20% reduction in power consumption while maintaining correction capability, and the ability to dynamically adapt decoding parameters for different types of block codes. The proposed architectural solutions demonstrate particular efficiency when processing long codewords (n > 1000), typical of modern 5G/6G communication systems and data storage systems

The article presents a technique for numerical modeling of the radiation pattern of an equidistant microphone array, taking into account the acoustic-mechanical interaction. We investigated the effect of mechanical deformations of the carrier board on the spatial characteristics of the array under the influence of a plane acoustic wave with a frequency of 500 Hz. We developed a model that includes a 260×260×2 mm board made of FR-4 material with 36 microphones arranged in 50 mm increments. We carried out comparative modeling for two cases: an elastic board and an absolutely rigid structure. We show that the maximum deformation of the board is 2.5993×10⁻⁹ m, and the average relative error of the common-mode sums of signals between the models does not exceed 0.0358%. To construct the radiation pattern, we used an algorithm for common-mode summation of signals, followed by normalization and conversion to a logarithmic scale. The results demonstrate that for a given configuration and frequency range, the effect of mechanical deformations on the spatial characteristics of the lattice is negligible. The results obtained make it possible to specify the limits of applicability of simplified models and develop effective algorithms for spatial signal processing without taking into account deformations of the supporting structure. The proposed technique can be used to optimize microphone array designs at the design stage.

The study addresses the features of using antenna-feeder devices (AFDs) with polymer coatings on unmanned aerial vehicles. The influence of polymer coatings on the electrodynamic characteristics of antenna-feeder devices operating in the frequency ranges of 915 MHz and 1200 MHz was investigated. Experimental measurements were carried out to assess the impact of polymer coatings made of glycol-modified polyethylene terephthalate (PETG), high-impact polystyrene (HIPS), polylactic acid (PLA), and acrylonitrile butadiene styrene (ABS). The dependencies of the resonant frequencies and reflection coefficients were determined. It was found that an increase in dielectric permittivity leads to a decrease in the resonant frequency and an increase in the reflection coefficient. The radiation efficiency is determined by the interaction of dielectric losses and the geometrical parameters of the polymer coatings. In the 915 MHz frequency range, the stability of the parameters is ensured by polymer coatings made of high-impact polystyrene (HIPS). The application of glycol-modified polyethylene terephthalate (PETG) coatings causes a shift in the resonant frequency of the antenna-feeder device. At 1200 MHz, stable matching is provided by polymer coatings made of high-impact polystyrene (HIPS) and white polylactic acid (PLA). The use of glycol-modified polyethylene terephthalate (PETG) results in increased reflection and a growth in the VSWR parameter. The obtained results can be applied in the design of antenna-feeder devices, taking into account the influence of the dielectric properties of polymer coatings.

The paper is devoted to effective methods of evaluating the OFDM (Orthogonal Frequency Division Multiplexing - multiuser version of digital modulation) communication channel of systems based on deep learning methods. In the context of growing requirements for the reliability and efficiency of wireless networks of 5G and Wi-Fi standards, an accurate assessment of the state of the communication channel is becoming particularly important. Traditional pilot-based methods are often not effective enough in rapidly changing signal transmission conditions. The paper proposes an adaptive algorithm for evaluating the characteristics of a communication channel based on the use of neural network models. The methodology includes the stages of selecting informative metrics (amplitude, phase), modeling channel operating conditions (signal-to-noise ratio (SNR), delay), and creating a dataset using the Rayleigh model. The architecture of a neural network combines precise and dense layers with an attentional mechanism to highlight the most significant features. The simulation results in the MATLAB/Simulink environment based on OFDM system models demonstrate the high efficiency of the proposed approach. The model achieves a coefficient of determination of 0.82 and a prediction accuracy of up to 0.95, surpassing traditional analogues. The key advantages are the high convergence rate, low computational costs, and the stability of the results when the input parameters change. The developed algorithm is a promising solution for implementation in modern communication systems to increase their noise immunity and adaptability.

The purpose of this article is to analyze the possibility of integrating memristors into the structure of headlamps, assess their impact on antenna characteristics, and demonstrate their advantages over traditional approaches. The relevance of this work is due to the need to overcome the limitations of classic headlights. Existing phase control technologies often require complex power circuits, have inertia, and limited reliability, especially in extreme conditions. Memristor solutions, on the other hand, offer high performance, low power consumption, and the ability to save state when power is turned off. Despite the growing interest in memristors in microelectronics, their use in antenna arrays remains poorly understood, which determines the scientific novelty of the research. The article is structured as follows: the first section provides an overview of the principles of operation of headlamps and key properties of memristors; the second section proposes the concept of a phased array antenna and describes the methods of its modeling. The third section is devoted to the analysis of the results and comparison with traditional systems; in conclusion, the prospects for the introduction of technology and possible directions for further research are discussed. The conducted research contributes to the development of adaptive antenna systems that combine high performance with miniaturization and energy conservation, which is especially in demand in satellite communications, the Internet of Things (IoT), and fifth-generation (5G) and Beyond mobile networks.

We show block diagram of the developed mobile acoustic signal detector. Due to the consideration of circular signal reception, we selected the number of microphones used in the solution, as well as their attachment points on the helmet base. We determined the optimal microphone model based on the criteria of sensitivity, frequency range; low level of intrinsic noise. We defined the model of the amplifying element, which ensures the matching of the microphone and the input of the amplifying cascade. We defined a model of an ADC that provides high-quality conversion of an analog signal into a digital one. We developed a circuit design solution for the device under development. We carried out the simulation of an analog signal reception unit at input signal levels of 100 MV, 10 mV, 100 mV, 500 mV. We obtained the frequency response of the amplifier and audio converter. According to the carrier and side harmonics levels at the output of the UHF model at an input signal level of 10 mV and a frequency of 1 kHz, we obtained a nonlinear distortion coefficient of 0.147%, which meets all the requirements for the system. We developed a model of an audio converter and carried out a simulation of the developed model.  We developed the topology of the printed circuit board, taking into account the fact that the system under development is mobile, the printed circuit board should have a low level of external interference. When designing the printed circuit board, we took into account the requirement of overlapping power lines and signal lines on the upper and lower layers. To prevent the appearance of phase shifts in the signals, we took into account the symmetry of the microphone arrangement.

The article discusses a successive approximation method, typical structure and blocks of an analog-to-digital converter based on a successive approximation register and a comparison of its main parameters with another type of ADC. We provide a description of the operation and interaction of the blocks of the designed register, main diagrams explaining the operation of the integrated circuit and possible non-standard ways of its application. A detailed representation of the blocks allows for a more complete understanding of the entire circuit's operation, identifying the principles of additional applications, and finding ways to improve its efficiency and operational schemes. We give analogs for the presented electrical circuits of the developed blocks. We present the main electrical characteristics and simulation results and give the layout view of designed chip. The integrated circuit is designed in standard 180 nm CMOS process (complementary metal–oxide–semiconductor). A specialized computer-aided design (CAD) system was used in the development of the electrical circuit and layout. The Spectre models provided by the silicon foundry were used for modeling the electrical circuit. The layout passed testing for compliance with design rules (Design rule check, DRC) and for conformity to the electrical circuit (Layout vs. Schematic, LVS).

We present a comprehensive design methodology for high-power switching power supplies (SPS) for special equipment used in computing, communication systems, medical equipment and electronic warfare, aimed at minimizing time costs and optimizing key characteristics. The methodology includes a step-by-step analysis of the technical task, selection of the optimal circuit topology (single-stroke or push-pull), calculation and modeling of the parameters of the power unit, in particular the transformer and chokes, using specialized software. Special attention is paid to the practical aspects of debugging prototypes. Using real-world examples, we analyzed typical problems that arise at the stage of creating a working sample: voltage surges during switching of key elements, leading to false alarms of the control system; parasitic resonance in the primary circuit of the transformer; increased output voltage ripples due to the failure of the supply chains. We proposed effective solutions and experimentally tested for each problem, such as correcting the switching time of transistors, shifting the operating frequency of the converter to avoid resonance, and selecting components resistant to high-frequency effects. We show that the application of the proposed technique makes it possible, therefore, to identify and eliminate design flaws, ensuring that the finished device meets the specified technical requirements.

The work is devoted to the development of an automatic optical control system for printed circuit boards (PCBs) of radio engineering devices based on deep neural networks. We considered the characteristic defects of PP that occur at the stages of photolithography, etching and metallization, and proposed methods for their automatic detection and classification. We show that using the YOLOv8 architecture (a computer vision model with segmentation of objects on images) in combination with image preprocessing algorithms makes it possible to achieve high accuracy in detecting defects with a processing time of less than 250 ms per frame. We researched and implemented strategies for accelerating inference using the OpenVINO and DeepSparse frameworks (open toolkits for developing computer vision and artificial intelligence applications), which increase productivity by 4-5 and 11 times, respectively. We developed a fault-tolerant distributed infrastructure based on Apache Kafka (a real-time messaging system between server applications) and PostgreSQL (an open source object-relational database management system), providing parallel data processing and integrity of results. We implemented the interface of interaction with the operator for notification of the status of processing and updating of models. An experimental evaluation of the system showed accuracy of up to 94 % and completeness of up to 89%. The application of the proposed methodology makes it possible to increase the yield of usable products, reduce control time and reduce operating costs in the production of printed circuit boards for radio engineering devices.

The article examines analytical and practical approaches to developing a riveting sequence for thin wing panels in order to improve assembly accuracy and ensure the required radial tension in high-resource joints using rivets OST 1-34012-76. The relevance of the study is due to the need to minimize residual deformations and stresses that may arise during riveting of thin-walled structures of aircraft equipment (AE), especially in conditions of limited technological space and high density of rivet rows with the creation of a rational stress-strain state (SSS). The article analyzes recommendations for existing riveting routes and options for preliminary fixation of elements, including the installation of process bolts and other temporary fasteners. The paper proposes an analytical method for selecting the location of stringers, panel configuration and features of the process equipment. Various schemes for processing stringers by sections, as well as the influence of the type of the type of process fasteners on the stability of the assembly geometry are considered. Experimental testing on the test wing panels showed that the use of a justified riveting sequence allows reducing the energy intensity of the assembly process by 30 – 40 %, ensuring the required radial interference of 1 – 2 %, respectively and maintaining dimensions within the tolerances established in the normative and technical documentation. The results obtained can be used in the design and organization of serial production of aircraft.

We considered technological issues of combined processing with the simultaneous use of several controlled physical and technical influences, including thermal, electrochemical, and mechanical processes of shaping the interface points of structural elements, using examples of multi-link power transmissions and long-length lines for transporting liquid working media. We show the influence of such facilities on the resource and the inter-repair period of aerospace and power engineering products at the stages of their creation, improvement and development in mass production. We justified the choice of typical research objects, taking into account the operating conditions of the products, and proposed a new evaluation indicator for this, the manufacturability criterion, which shows the quantitative level of efficiency of the technological process option under consideration. Thus, we scientifically substantiated the legitimacy of the choice, appointment or development, creation of new technological solutions that determine and develop the basic indicators in the product and ensure an increase in its quality as a whole. On this basis, we developed promising technological techniques and means of equipping production, which made it possible to build a mechanism for the formation of combined processes with the expanded use of accumulated industry experience in the successful application of physical and technical impacts, including those protected by state security documents. In the article, we gave examples of the use of inventions, most of which were obtained by scientists of the VSTU scientific school, including the author's invention and his articles. They are adapted to the operating conditions of transmission and main gears used in aerospace and energy engineering, which allowed the author to become one of the performers of programs at the industry and state levels.

The article presents the results of a comprehensive gas dynamic analysis of an unmanned aerial vehicle (UAV) of the tiltrotor type using computational fluid dynamics (CFD) methods in the Ansys Fluent environment. A special feature of the gas dynamic analysis is the purge model of the aircraft with two multicopters (payload) placed on its surface, which significantly reduce the aerodynamic characteristics of the UAV. The relevance of the work is due to the growing demands on the aerodynamic efficiency and controllability of vertical take-off and landing UAVs, especially in transient flight modes. In contrast to existing studies focusing on isolated modes, this paper provides a comparative analysis of aerodynamic characteristics (pressure fields, speeds, lift, drag) for five critical angles of attack.: -2°, 0°, 2°, 4° and 8°. Based on the data obtained, the key zones of formation of aerodynamic moments and drag were identified, which made it possible to formulate recommendations for optimizing the geometry of the bearing surfaces and the layout of the apparatus. The analysis demonstrates not only the correctness of the calculation method used, but also reveals the specific features of the flow around a tiltrotor with a payload placed on board, which must be taken into account to increase stability and energy efficiency in aerodynamically unfavorable flight conditions. The results of the work are of practical value for the design stage of UAVs of this class, allowing to reduce the number of field experiments and purposefully optimize the design.