Science-based technologies

MATHEMATICAL MODEL OF TRANSMISSION OF STOCHASTIC SIGNALS UNDER INTERFERENCE CONDITIONS

Maksim Gariachiy , Serhii Shcherbinin — Thu, 28 May 2026 00:00:00 +0300

This article is devoted a mathematical model of stochastic signal transmission under conditions of radio-electronic interference based on the combined application of the Karhunen–Loev transform and adaptive Volterra series. The relevance of the study is due to modern requirements for combat control information systems, which in the conditions of large-scale armed aggression against Ukraine have become the main means of operational, tactical and strategic control of troops. Traditional linear signal processing methods do not provide the necessary compromise between spectral compactness, noise immunity and adaptability under conditions of non-stationary interference, multi-path fading and active radio-electronic interference, and also do not fully take into account the stochastic nature of signals, channel memory effects and nonlinear distortions arising under the influence of interference. To solve the problem, a combined application of the Karhunen–Loev transform for preliminary decorrelation and energy densification of stochastic signals with subsequent adaptive nonlinear processing using second-order Volterra series is proposed. The adaptability of the system is ensured by a multi-criteria optimization functional with dynamic adjustment of weight coefficients depending on the combat application scenario (stealth mode or active electronic warfare). The model describes the process of signal transmission through the channel, taking into account the impulse response of the environment and the generalized interference effect. The results of the study were obtained by simulation modeling in the Python 3.11 environment. A comparative analysis of the proposed approach with traditional signal processing methods (Fourier filtering, wavelet transform, Z-transform, classical Volterra model and separate application of the Karhunen–Loev transform) for different types of stochastic signals is carried out. The proposed approach provides increased spectral efficiency, noise immunity and adaptability of combat control information systems, which creates the prerequisites for its practical implementation in wireless combat control systems, secure communication facilities and sensor networks, where they are of particular importance.

METHOD OF SYSTEMATIC CHANNEL-BLOCK PERMUTATIONS IN A QOFDM SIGNAL ENSEMBLE

Ivan Kharchenko , Volodymyr Lysechko — Thu, 28 May 2026 00:00:00 +0300

The article presents QOFDM-SP (Subcarrier Permutation) as a further development of quasiorthogonal frequency access on subcarrier frequencies. The method does not introduce a new physical-layer signal and does not change the frequency plans, subcarrier spacing, subcarrier positions, or threshold admissibility condition of the base QOFDM ensemble.

The proposed method adds a systematic channel-block assignment layer over already defined frequency plans. In this layer, each frequency plan is divided into logical channel blocks, and the assignment of channels to these blocks is varied between plans by permutations. A reproducible structural experiment is performed for an ensemble with K = 24, L = 4, ΔF = 20 MHz. The selected values of the number of subcarriers are the first 24 prime integers not smaller than 23.

Since these values are pairwise coprime, each pair of plans has one endpoint coincidence at the upper edge of the common band, and the maximum same-channel coincidence count remains equal to one for all tested strategies. The main structural effect is therefore observed in the accumulated number of same-channel coincidences and in the number of plan pairs with such coincidences: both the cyclic and all-factorial assignments reduce these values from 276 to 60 and from 276/276 to 60/276, respectively.

Thus, within the adopted discrete structural model, the proposed QOFDM-SP method provides a more uniform channel-block organization of QOFDM signal ensembles and reduces the concentration of same-channel frequency-position coincidences without modifying the base QOFDM frequency plans.

METHOD FOR DETECTING ANOMALOUS LOCAL UPDATES AND ISOLATING MALICIOUS PARTICIPANTS IN FEDERATED LEARNING SYSTEMS

Stanislava Kudrenko , Oleksii Nimych, Ihor Makieiev — Thu, 28 May 2026 00:00:00 +0300

This paper proposes a method for detecting anomalous local updates and isolating malicious participants in federated learning systems. The method is aimed at improving the resilience of distributed machine learning models to model poisoning attacks and distorted local updates. The proposed approach combines two protection levels: anomaly assessment of local updates based on their deviation from the collective update pattern and isolation of participants that repeatedly demonstrate suspicious behavior. An anomaly score is used to evaluate local updates, while an accumulated anomaly counter is applied to control long-term participant behavior. The paper presents the formalization of the proposed method, describes the adaptive anomaly threshold mechanism, and defines the procedure for forming a trusted participant set for further global model aggregation. An illustrative numerical example is provided to demonstrate the operation of the proposed approach and to show the possibility of detecting anomalous local updates while reducing their influence on the global model. A comparative analysis of the proposed approach with existing aggregation and malicious participant detection methods in federated learning systems is also presented. It is shown that, unlike conventional aggregation schemes, the proposed method provides not only suspicious update filtering but also control of repeated anomalous participant behavior. The proposed approach can be applied in federated learning systems for edge and fog environments, as well as in information systems of critical infrastructure facilities.

METHOD FOR THE HIERARCHICAL DYNAMIC SELF-ORGANISATION OF COMPUTING NODES IN SELF-ORGANISING TELECOMMUNICATIONS SYSTEMS

Olena Lozko , Volodymyr Pastushenko — Thu, 28 May 2026 00:00:00 +0300

У статті розроблено метод ієрархічної динамічної самоорганізації обчислювальних вузлів у самоорганізованих телекомунікаційних системах, орієнтований на функціонування в умовах змінної топології, неоднорідності ресурсів, динамічного трафіку та часткових відмов. Запропонований підхід базується на формалізації системи у вигляді динамічного зваженого графа та передбачає побудову багаторівневої структури управління з виділенням локальних кластерів, координаторів кластерів і вищих рівнів агрегування. У межах методу визначено функцію близькості для кластеризації вузлів, інтегральний рейтинг вибору координаторів, правила резервування та подієву логіку реконфігурації внутрішньокластерних і міжкластерних зв’язків. На відміну від підходів, у яких кластеризація, вибір координаторів і перебудова структури розглядаються окремо, у роботі ці компоненти інтегровано в єдиний механізм керованої багаторівневої самоорганізації. Обґрунтовано систему критеріїв оцінювання, що включає середню затримку, нерівномірність завантаження, безперервність обслуговування, час реконфігурації та інтегральний показник живучості. Сформовано структуру імітаційного експерименту для подальшої верифікації запропонованого методу та визначено формат порівняльного подання результатів. Практична цінність роботи полягає у можливості застосування методу в edge/fog-інфраструктурах, розподілених обчислювальних платформах і спеціалізованих телекомунікаційних мережах, де критичними є стійкість, масштабованість і оперативність реконфігурації.

THEORETICAL FOUNDATIONS OF EQUIVALENT TRANSFORMATION OF UNEQUAL-ENERGY COMPLEX SIGNAL ENSEMBLES FOR CODE DIVISION MULTIPLE ACCESS SYSTEMS

Oleksandr Zhuchenko , Halyna Shubina — Thu, 28 May 2026 00:00:00 +0300

A promising direction in the development of ultra-wideband access systems is the transition to code division multiple access based on ensembles of unequal-energy complex pulse signals. In contrast to equal-energy signal ensembles, for which the relations for efficiency evaluation mostly have a simple form suitable for application, the efficiency evaluation of unequal-energy complex signal ensembles usually requires the use of the corresponding relations in general forms. The article develops the theoretical foundations of the equivalent transformation of unequal-energy complex signal ensembles for code division multiple access systems, which provide the reduction of the relations for efficiency evaluation to a form similar to that of the equal-energy case. An approximate transformation of an unequal-energy complex signal ensemble into an equivalent equal-energy complex signal ensemble, equivalent in terms of the total signal energy, is determined. This transformation represents the effect of nonuniform signal energy distribution on multiple access interference through a change in the volume of the equivalent equal-energy complex signal ensemble at a constant value of the maximum ensemble cross-correlation. This is achieved by the approximate reduction of the set of signal energies to a scalar value of the volume of the equivalent equal-energy complex signal ensemble, which results in reducing the relations for efficiency evaluation of unequal-energy complex signal ensembles to a form similar to that of equal-energy complex signal ensembles. To implement the equivalent transformation, two interrelated variants of approximate scalar reduction of the set of signal energies of the ensemble to a scalar value are proposed. These variants are based on a concentration measure and its inverse concentration normalized to the equal-energy value. They have opposite effects on the change in the volume of the equivalent equal-energy complex signal ensemble and thereby define two different interpretations of the effect of multiple access interference on users with the lowest and highest signal energies, which directly corresponds to different degrees of manifestation of the signal dominance effect.

METHOD FOR DETERMINING THE INFORMATIVENESS OF SEGMENTS FOR INTELLI-GENT VIDEO CONTENT PROCESSING SYSTEMS

Vladimir Barannik, Pavlo Pertsev — Thu, 28 May 2026 00:00:00 +0300

The article shows that currently a number of effective modern video data encoding technologies have been created, which are based on standardized video frame compression formats and dynamic streams. These include: H265/HEVC and H266/VVC. Accordingly, a number of directions are outlined for taking into account the features (complexity) of the content for these technologies. The first type of approaches is based on the principle of reducing the bitrate while complying with the requirements for the level of root-mean-square indicators, which determine the level of integrity. The disadvantages here are: increasing the complexity of processing due to the multi-iterative process of constructing a syntax tree; the possibility of making only a rough estimate of the bitrate level before the compression process (entropy coding) begins. The second direction is based on approaches based on the use of artificial intelligence models. At the same time, this approach has certain limitations. This concerns the increase in the complexity of processing and time delays in the process of learning and forming a latent space. In addition, errors occur in the process of identifying objects. One of the effective directions for increasing the efficiency of semantic selection of video segments in the compression process is the development of combined approaches. In this case, the basic one is the identification (classification) of segments by the level of semantic complexity on the basis of a set of features that are detected at the level of syntactic perception of video information. The main task here is the selection of such features that will simultaneously have an associative dependence with the level of semantic complexity of video segments and the permissible energy efficiency of the computational process. Methodological foundations have been developed for the threshold-metric classification of standardized segments by size into two basic classes based on the scaling of the results of division into classes of its local segments. The basic ones here are: squaring the segment by local segments with the formation of an information model for each of them by a set of structural and statistical parameters; Scaling to establish the level of complexity of a video image segment from the perspective of the potential of its features with respect to the presence of redundancy is carried out on the basis of a decision rule based on the analysis of frequency information of the results of dividing local segments into two basic classes; determination of the complexity class of local segments based on relating the metric to one of two permissible threshold intervals with a defined limit.

METHODOLOGY FOR ASSESSING THE RELIABILITY OF MONITORING THE RELIABILITY STATE OF A BASIC ELEMENT OF A TELECOMMUNICATIONS NETWORK

Bohdan Zalevskyi — Thu, 28 May 2026 00:00:00 +0300

At the present stage of development of social relations and their important component—economic interconnections—the problem of ensuring the reliability of telecommunications networks (TCNs) in the process of data transmission is becoming increasingly significant and relevant. Comprehensive consideration of all factors affecting the reliability level of a basic element sharply complicates monitoring models and requires the development of new approaches and methods for their analysis.

This paper develops a methodology for assessing the reliability (trustworthiness) of monitoring the reliability state of a basic element of a telecommunications network. The proposed methodology makes it possible to ensure the reliability of monitoring the reliability state in a multi-level monitoring system of TCN basic elements according to the criterion of the lower bound of the conditional probability of correct classification when monitoring a basic element at a specified level.

It is shown that the lower bound of the conditional probability of correct classification when monitoring a basic element depends on the number of levels in the alphabet and is determined according to the decision-making criterion based on the maximum a posteriori probability method.

It is established that the conditional probability of correct classification when monitoring a basic element depends on the magnitude of the error; an increase in the error leads to an increase in the values of the conditional probability of correct classification, i.e., the reliability of monitoring the reliability state of a basic element of a telecommunications network.

RESONANT FREQUENCY DISTRIBUTION OF NON-UNIFORM LINES

Valeriy Kozlovsky — Thu, 28 May 2026 00:00:00 +0300

The article defines the conditions that must be satisfied by the resonant frequencies (spectrum) of non-uniform transmission lines loaded with a reactive lumped load consisting of a finite number of lumped inductances and capacitances. To determine the conditions for the physical realizability of non-uniform lines, the theory of continued fractions is applied. In this approach, the input impedance of the loaded non-uniform line is represented as an infinite ladder network, obtained through a limit transition from a finite continued fraction to an infinite one.

To determine the line delay time and fulfill the physical realizability conditions, a Richards' rod circuit is utilized (a multi-stage line consisting of a cascade connection of uniform transmission line segments with different characteristic impedances and equal delay times for all stages). A methodology has been developed to determine the values of the lumped load elements, allowing the elements of the continued fraction to be expressed through the frequency spectrum of the loaded line. The core of this methodology lies in the fact that the reactive load can be represented as the last element of a ladder network comprising both the line and the load. Consequently, to determine the load elements, a ladder network is first constructed, and the impedances of the final elements are expressed via the line spectrum. To transition to a loaded long line, a limit transition to an infinite ladder network is performed. This approach has yielded analytical expressions in the form of series for determining the elements of the lumped load circuit. Based on the obtained relations, the construction of convergents for lossless distributed circuits is considered. The results enable the synthesis of resonant systems with a required distribution of resonant frequencies.

Significant attention is paid to the deformation of resonant frequency spectra: conditions are established under which new resonant frequencies can be added or existing ones excluded within a limited frequency range. Furthermore, the variation of load elements is studied, demonstrating that spectrum deformation leads not only to changes in the values of the continued fraction elements but also to a change in the total number of load elements.

A METHOD FOR IMPROVING SPEECH RECOGNITION EFFICIENCY BASED ON GENETIC OPTIMIZATION OF WAVELET FUNCTIONS

Oleksandr Lavrynenko — Thu, 28 May 2026 00:00:00 +0300

This article addresses the pressing issue of ensuring high reliability in the operation of speech recognition systems under conditions of noise interference. The scientific novelty of this research lies in the development of a method for synthesizing an optimal adaptive wavelet kernel for the initial layers of convolutional neural networks. Unlike existing approaches, which rely on stochastic weight initialization or the use of strictly deterministic basis functions (such as Meyer, Daubechies, or Simlet wavelets), the authors propose an algorithm for the targeted formation of kernel geometry based on Akima interpolation splines. The central focus of the study is the process of optimizing wavelet morphology, where the mean-squared error of the discrepancy between the amplitude-frequency response of the synthesized filter and the energy spectral portrait of a specific speech signal is chosen as the objective function. To solve the problem of minimizing this function in the multidimensional space of spline parameters, a modified parallel genetic algorithm is applied. The use of evolutionary search allows for effectively overcoming the problem of local extrema, characteristic of non-convex surfaces of objective functions, when searching for optimal ordinates of spline nodal points. The article provides a detailed analysis of the algorithm’s convergence up to the 50th generation and an assessment of computational efficiency depending on the number of processor cores used. The results of comparative modeling are presented, confirming the superiority of adaptive kernels over classical analytical wavelets. In particular, the implementation of an optimal adaptive filter into the structure of a convolutional classifier allowed for an increase in speech recognition accuracy by 15–22% at low signal-to-noise ratios of 5–15 dB. It is shown that, thanks to the use of parallel computing schemes, the system’s adaptation time to a new speaker is reduced to 2.1 seconds, which opens up broad prospects for integrating the method into robust voice control systems for unmanned robotic systems and specialized information and communication networks.

INFORMATION TECHNOLOGY OF MOBILE TOMOSYNTHESIS WITH BONE DENSITY ASSESSMENT

Thu, 28 May 2026 00:00:00 +0300

The paper considers approaches to improving the efficiency of mobile X-ray tomosynthesis under conditions of limited energy resources and mechanical instability of the system. It is shown that, compared to stationary systems, mobile X-ray complexes operate with significantly lower source power and restricted exposure parameters, which complicates the achievement of high image quality and requires optimization of acquisition and processing procedures in accordance with the ALARA principle. The relationship between the signal-to-noise ratio of the primary X-ray quantum flux and the signal-to-background ratio in reconstructed images is analyzed. It is demonstrated that reducing the tomographic layer thickness, inherent to tomosynthesis, decreases the effect of anatomical structure superposition and increases image contrast. However, this also leads to a reduction in the number of detected quanta per voxel, resulting in a lower signal-to-noise ratio. A compromise approach is substantiated, involving controlled reduction of spatial resolution in the image plane to stabilize the signal-to-noise ratio without increasing radiation dose. A method for compensating mechanical and geometric errors of a mobile system is proposed. The method is based on determining the actual trajectory of the X-ray source using a reference marker and inertial sensor data. This allows the formation of individual geometric parameters for each projection and enables adaptation of reconstruction algorithms to real acquisition conditions, reducing artifacts and improving image quality. In addition, a method for bone density estimation based on intensity calibration using a reference phantom is developed. The proposed approach enables relative quantitative assessment without requiring full 3D reconstruction and provides consistency of measurements across different studies. The results demonstrate that the integration of the proposed methods improves image quality and enhances the functional capabilities of mobile X-ray tomosynthesis systems in practical applications.

METHODS FOR MONITORING THE CONDITION OF AIRCRAFT ELECTRONIC SYS-TEMS USING THE INTERNET OF THINGS

Anton Plugoviy — Thu, 28 May 2026 00:00:00 +0300

The article considers methods for monitoring the condition of aircraft electronic systems using Internet of Things technologies. It is shown that IoT monitoring makes it possible to move from a single check of individual electronic units to the systematic collection, transmission, accumulation and analytical processing of diagnostic data during aircraft operation. Special attention is paid to the operation of an aging aircraft fleet, for which the individualization of maintenance, consideration of the actual operational history of a specific aircraft and timely detection of hidden degradation processes are of particular importance. It is substantiated that diagnostic parameters of electronic systems have a stochastic and non-stationary nature, while their trends may change under the influence of flight modes, temperature loads, vibration, power supply quality, maintenance actions and other operational factors. The need to take into account the change-point effect is considered, when the statistical properties of the observed process change and further forecasting should be performed for a new quasi-stationary stage. A generalized scheme of IoT monitoring data flow from the aircraft to the ground-based technical operation system is proposed. This scheme includes the aircraft maintenance base, maintenance hangar, engineering center, reliability service, operational database and decision support system. The article considers methods for processing monitoring data, including threshold analysis, trend analysis, event analysis, statistical estimation, regression analysis, statistical classification, anomaly detection, forecasting and data processing under the change-point effect. The use of data quality indicators, an integral monitoring indicator, a maintenance priority indicator and criteria for selecting diagnostically significant parameters is substantiated. The practical significance of the work lies in the possibility of improving diagnostic reliability, reducing troubleshooting time, decreasing the number of unjustified replacements of electronic units, improving maintenance planning and increasing the economic efficiency of aircraft operation.

METHOD OF STEGANOGRAPHIC HIDING IN AUDIO CONTAINERS USING LLMS

Thu, 28 May 2026 00:00:00 +0300

Background. The development of digital networks requires the improvement of methods for covert data transmission, among which the LSB approach is the most common due to its simplicity and high capacity. However, modern stegoanalysis methods necessitate a deep systematization and critical analysis of existing research to increase the stability and adaptability of algorithms. The goal is a comprehensive analysis and systematization of modern research in the field of LSB steganography to identify current trends and form an analytical base that will allow improving methods for covert information transmission. Methods. The method of systematic analysis and classification of modern scientific publications in the field of digital data protection was used. The study covered the analysis of the evolution of LSB (Least Significant Bit) methods, in particular their hybridization with cryptographic algorithms (AES, RSA), frequency transforms (DWT, DCT) and stochastic approaches (chaotic maps). Special attention was paid to the use of Reed-Solomon coding, adaptive pixel selection and machine learning models to optimize data embedding. Results. An adaptive audio steganography method has been developed that uses large language models (LLM) to dynamically control the depth of LSB embedding ($k$) based on psychoacoustic analysis and local signal entropy. Experimental simulation in MATLAB confirmed the superiority of the method over classical analogues, in particular, the highest peak signal-to-noise ratio (PSNR = 58.23 dB) and the minimum mean square error (MSE = 0.098) were achieved. Conclusions. The integration of LLM into the steganographic coding process allows us to transform the LSB approach into an intelligent system that adapts to the semantic context of the audio container, providing high visual and statistical invisibility. Despite the increase in processing time, the proposed algorithm is optimal for secure data transmission systems due to its ability to minimize the impact on the human auditory system (HAS) and resist modern stegoanalysis methods.