Problems of Informatization and Management

Software architecture of the system for expert assessment of the efficiency of unmanned aviation systems

2025-12-05T12:26:15+02:00

This paper proposes a software architecture of an automated system for expert assessment of the quality of unmanned aviation systems, designed to collect individual expert assessments, their statistical processing and generation of interactive reports and recommendations for improving the ergonomic characteristics of UAS. The program allows you to systematize expert assessment of the quality of unmanned aviation systems when direct measurements are impossible or ineffective. The program is implemented on the basis of a cross-platform Web solution with a client part in HTML5/JavaScript and server processing in PHP, the modular structure of the system is given (user account, project creation, expert evaluation module, data analysis, reports and recommendations), approaches to data storage and protection, authentication mechanisms and an algorithm for aggregating results based on a weighted average are described. Functional capabilities for dynamically adjusting weights and thresholds are proposed, as well as tools for ensuring the integrity and accessibility of examination results. The final part of the work provides an example of software implementation and its interface.

Comparative analysis of convolutional and multilayer perceptron neural networks for resource allocation in distributed computing systems

2025-12-05T11:58:23+02:00

Efficient resource distribution in heterogeneous distributed computing systems requires intelligent node selection mechanisms capable of adapting to dynamic system conditions. This research presents a comparative evaluation of Convolutional Neural Network (CNN) and Multilayer Perceptron (MLP) architectures for binary node suitability classification in distributed task scheduling environments. The study employs five synthetic datasets ranging from 100 to 2000 instances, with each node characterized by four critical attributes: performance, security level, baud rate, and reliability. Experimental results demonstrate that the MLP architecture achieves validation accuracy between 91% and 98.8% with high consistency across dataset sizes, while the CNN architecture shows fluctuating performance ranging from 85% to 94%. The key finding reveals that MLP architectures outperform CNNs for tabular node data due to better alignment with unstructured attribute relationships, as fully connected layers naturally handle unordered features without imposing spatial locality assumptions. This empirical analysis provides practical guidance for architecture selection in AI-based resource schedulers, demonstrating that simpler architectures can outperform more complex ones when appropriately matched to problem structure. The findings contribute evidence-based recommendations for distributed systems practitioners implementing neural network-based scheduling solutions.

Intelligent load balancing management in cloud web hosting: Evaluation criteria and methodology

2025-12-05T01:11:39+02:00

This paper presents a methodology for selecting a software load balancer designed for cloud-cluster web hosting environments that use a large number of Transport Layer Security (TLS) certificates, perform their dynamic rotation, and integrate with an asynchronous network traffic analysis system. It is shown that traditional methods of evaluating load balancers, based on static testing of throughput and response time, do not reflect real operating conditions, where the system configuration constantly changes, block lists are updated, the composition of servers varies, and compliance with the Service Level Agreement (SLA) is required. Under such conditions, the load balancer should be considered an active control element capable of handling tens of thousands of certificates, applying configuration changes without process restarts, updating internal in-memory data structures, and providing reliable telemetry for automatic scaling systems. A new evaluation approach is proposed, based on the concept of a “dynamic benchmark,” which combines load testing with gradual configuration complexity: increasing the number of domains, updating certificates, applying new block list entries, and changing cluster topology. Key evaluation criteria are defined, including scalability, resilience under dynamic changes, flexibility for implementing custom logic, manageability, and completeness of observability. The developed methodology eliminates the existing gap between academic research and the practical operation of modern hosting platforms and creates a foundation for further experimental validation and the development of a next-generation load balancer prototype focused on reliability, adaptability, and intelligent load management.

Information technologies for optimization of expert systems and real-time databases

2025-12-05T00:49:54+02:00

The purpose of the article is to present an integrated approach to the priorities of key performance parameters (KPIs) of the information and communication network in terms of the criteria of specificity, measurability, viability, and realistic goal setting.
Planning/methodology/approach – the study was built using the Analytical Hierarchy Process (AHP) as the basis for pairwise comparisons with consideration of each KPI.
Novelty – a new approach was developed aimed at step-by-step directions for making decisions for the KPI prioritization process. The results of the analysis of this problem highlight the possibility of applying the proposed approach and the calculation process for KPI prioritization.
Research limitations/applicability – Rating scales in the AHP are conceptual; the research was conducted at the departments of expert systems, databases, and computer information technologies
Originality/value – The article presents a new approach to KPI optimization. The proposed approach is a holistic mechanism; it will be able to provide teams with decision-making; it is also able to improve the quality of advanced technological approaches; finally, it provides broad opportunities for future research.

Method of accelerated modular exponentiation based on dynamic pre-calculation tables

2025-12-04T23:05:24+02:00

A method for accelerated calculation of the modular exponent has been theoretically substantiated, developed and investigated, the difference of which is that in each iteration, recalculations are performed, which allow performing modular operations of multiplication and squaring with group processing of bits in both phases: multiplication itself and Montgomery reduction, due to which the time for implementing exponentiation is reduced.
Acceleration is achieved by forming recalculation tables at each iteration of the expo-nentiation algorithm. The paper theoretically substantiates this possibility by using the built-in RAM of the terminal device, provides a detailed description of the method, the operation of which is illustrated by a numerical example, and presents a detailed assessment of its effectiveness.
It is theoretically and experimentally shown that the proposed method allows for 5 times faster modular exponentiation compared to known methods for its implementation on terminal devices in IoT technologies.

Mathematical modeling of an automated monitoring and predicting system of the psychophysiological states of a navigator under the influence of fatigue and stress

2025-12-04T22:36:19+02:00

A ship is a complex hierarchical human-machine control system (ergatic system) in which a person is a key component responsible for decision-making at all levels, and therefore the negative effects of the human factor, due to fatigue and stress, on the safety of control are possible at any of these levels, regardless of the operator's qualifications or experience.
Modern automated ship control systems are focused primarily on technical movement parameters and navigation conditions, but hardly take into account the psychophysiological state of operators. This creates additional risks for maritime safety, as stress and fatigue significantly affect the navigator's performance and ability to make the right decisions in difficult situations.
The object of the study is the process of automated monitoring and control of changes in the navigator's psychophysiological parameters under the influence of stress and fatigue during the performance of navigation watch duties.
The problem lies in the lack of effective technical solutions that would ensure the timely detection and compensation of the negative effects of stress and fatigue, which are not taken into account by modern automated navigation systems.
The study proposes the creation of an automated module for monitoring the navigator's condition, based on multiplicative modeling of time series of physiological parameters, taking into account circadian rhythms, adaptive dynamics, and fatigue factors. To improve accuracy, robust regression and transfer function analysis methods are used in the human-machine system.
The scientific novelty lies in the application of a conceptual model that decomposes physiological factors into influence coefficients and allows modeling and predicting the navigator's reactions in real time.
Experimental tests conducted using the Navi Trainer 5000 system and on a real ship confirmed the effectiveness of the model: the accuracy of the assessment exceeded 90%.
The practical significance of the results obtained lies in the possibility of integrating the model into the intelligent modules of ship control systems. This provides automatic adjustment of watch schedules, dynamic adaptation of autopilot parameters, and the formation of warning signals for the crew, which increases the level of maritime safety.

Comparative analysis of monocular depth estimation in satellite two-dimensional images

2025-12-04T17:19:14+02:00

With the rapid growth in the use of satellite mapping services, there is a need for simple and effective methods and means of constructing 3D (three-dimensional) reconstructions of terrain using only 2D (two-dimensional) images as input, which ensure the reproducibility and comparability of results in applied urban analysis and navigation tasks. Monocular depth prediction models are a tool for formalizing such processes, as they allow the relative structure of a scene to be reconstructed with a minimum amount of input data, which helps to assess the relief and height contrasts of objects. Despite the availability of services that allow obtaining three-dimensional satellite images, their limitation is that they are either commercial or do not work in Ukraine. That is why the task is set to build such a model that will solve these problems. The paper considers pre-trained models and their application to satellite images. The models considered were trained on different datasets, so the quality of reconstruction was evaluated on pre-trained models, selected by quality, and trained on a dataset of satellite images. The results of the study can be used to develop methods for aircraft navigation, reconstruction, and terrain analysis, and allow for simpler and faster 3D reconstruction of arbitrary objects, although they are inferior in accuracy to resource-intensive methods. It should be noted that the proposed methodology allows the results to be applied to applied problems in other domains. It is reproducible, easy to use, and suitable for further scaling to larger samples and metric calibration problems.

Overview of the hardware implementation of devices using Posit number system

2025-12-04T16:58:26+02:00

The modern development of computer computing is accompanied by increasing demands for numerical representation accuracy, energy efficiency, and hardware optimization. The traditional floating-point format IEEE 754, which has served as the fundamental standard in numerical computations for decades, reveals a number of limitations that become critical under new technological challenges. In particular, issues such as excessive hardware complexity, limited accuracy, and inefficient utilization of computational resources remain problematic.
The Posit arithmetic is considered a promising alternative to classical approaches. Its conceptual features make it possible to enhance numerical accuracy, reduce hardware costs, and achieve a better balance between performance and energy efficiency. This makes the Posit system a relevant subject of investigation in such fields as deep learning, embedded systems, and high-performance computing.
However, the practical implementation of Posit arithmetic at the hardware level requires in-depth analysis. The viability and scalability of this new number representation system are determined by the efficiency of hardware architectures capable of providing the required levels of accuracy, performance, and robustness against computational errors.
In the main part of this article, prepared within the framework of obtaining the degree of Doctor of Philosophy, the fundamentals of Posit arithmetic, its structure, advantages, and challenges in comparison with IEEE 754 are described.This is followed by a detailed examination of hardware implementations of Posit arithmetic units. Applications in deep learning, accuracy and efficiency analyses, robustness studies, and error investigations will be considered. The review concludes with a summary of the key findings and implications for future research.

Intelligent approaches to designing cloud-oriented decision-support systems in education

2025-12-04T15:54:23+02:00

The article presents a comprehensive approach to designing an intelligent cloud-oriented decision-support system (DSS) for the educational domain, combining the principles of microservice architecture, cognitive modelling, and machine-learning operations. The relevance of the study is justified by the growing volume of educational data, the need for personalized learning trajectories, and the necessity to maintain stable SLO indicators under dynamically changing learning conditions. The analysis of current research highlights the limitations of traditional monolithic LMS platforms, which lack sufficient flexibility, scalability, and capabilities for integrating intelligent models.
The proposed system architecture is based on event-driven microservice interaction via Kafka and incorporates modules for collecting and normalizing learning events, a feature-engineering subsystem, recommendation and risk-prediction services, and a dedicated pipeline for modelling and monitoring ML components. The results of the study demonstrate that combining microservice decomposition with intelligent data-analysis methods improves recommendation accuracy, enhances performance indicators, and ensures the resilience of the educational platform under high load. The presented architecture can serve as a foundation for building scalable and adaptive next-generation educational ecosystems.

Event-driven edge processing model for industrial IoT systems

2025-12-04T11:26:10+02:00

Efficient and reliable cold chain management requires rapid detection and interpretation of temperature deviations, equipment anomalies, and operational disturbances. Traditional IoT-based monitoring architectures rely heavily on cloud connectivity, resulting in latency, reduced autonomy, and limited resilience in time-critical situations. This paper proposes an event-driven processing logic integrated into an edge computing architecture for intelligent cold chain logistics platforms.
The model introduces a structured four-tuple event representation, a deviation descriptor for quantifying abnormal behaviour, an adaptive severity classification function, and a rule-based edge decision mechanism. Formulas (1)–(4) formalize the transformation of raw sensor measurements into contextualized events, analytically evaluated deviations, severity levels, and real-time edge-level actions.
Modelling results using realistic temperature trajectories—stable control, slow thermal drift, and door-abuse patterns—demonstrate the model’s ability to distinguish benign fluctuations from harmful deviations, escalate severity dynamically, and operate autonomously during connectivity interruptions. The proposed approach significantly enhances responsiveness, resilience, and operational reliability of cold chain monitoring systems.

Evaluation of the possibilities of creating interference-resistant codes by selecting code word combinations

2025-11-25T18:26:46+02:00

The work is devoted to the issues of finding the best codes capable of correcting the maximum possible number of errors for a given dimension. It is proposed to carry out such work by searching for possible combinations of code words using high-performance computing equipment. The main theoretical provisions are substantiated. Full and simplified algorithms are proposed that implement the method of building up, which allows improving the clarity and to significantly increase the speed of the algorithm, while losing accuracy not worse than the unit value of the maximum code distance. Approximate estimates of the memory and computation time required for the execution of the algorithm are given, which will be useful during the further implementation of the algorithm on high-performance computing devices. An upper estimate of the required computational resources of two main types is found – the amount of memory and machine time, as functions of the dimensionality parameters of the noise-resistant code.

Improving the reliability of synchronous electric motors through monitoring of excitation parameters and winding temperature

2025-11-24T18:29:06+02:00

The article examines the influence of excitation parameters and thermal conditions on the reliability of synchronous electric motors. Analytical dependencies are developed to describe the variation of the electromagnetic torque as a function of the excitation current, as well as thermal models that take into account heat losses in the windings. An integrated mathematical model is proposed for evaluating the failure rate of the motor, considering the combined effect of electrical and thermal factors. A monitoring system approach is presented, providing real-time control of excitation and temperature parameters to enable timely detection of pre-failure operating modes. The obtained results can be applied to improve the durability, energy efficiency, and fault-free operation of synchronous machines in industrial electric drives.

Load balancing and cloud node resilience enhancement algorithm based on predicted integral index

2025-11-24T18:16:06+02:00

The paper addresses the problem of load balancing in cloud environments designed for processing stochastic IoT traffic. It is established that traditional reactive methods are insufficiently effective under conditions of sharp fluctuations in request intensity. An adaptive routing algorithm based on the predicted integral node load index is proposed. This index aggregates CPU, RAM, disk I/O, and network activity metrics into a single criterion. An LSTM recurrent neural network model is used to forecast node states. The problem of minimizing expected service time and overload is formalized. The implementation of the proposed approach enables preventive resource distribution, thereby enhancing system stability. Furthermore, the proposed approach significantly contributes to the cybersecurity and resilience of cloud infrastructure. Within IoT ecosystems, node saturation frequently leads to compromised service availability and heightened susceptibility to Denial-of-Service (DoS) attacks. Leveraging the predicted integral load index facilitates not only the optimized allocation of computational resources but also the early detection of load anomalies attributed to aberrant IoT device behavior or hostile actions. This fosters the integration of load balancing with security monitoring frameworks, bolstering cloud fault tolerance and minimizing service degradation risks arising from both systemic failures and malicious intent.

Machine learning-based anomaly detection with Isolation Forest in large-scale data analysis

2025-11-24T18:03:39+02:00

This paper presents an applied study of unsupervised anomaly detection with Isolation Forest on large multivariate sensor data. The study implements a concise Python workflow that acquires city‑scale measurements for Kyiv, merges reference metadata, removes invalid records, selects two continuous indicators, and trains Isolation Forest with parameters. Temporal analysis shows that anomalies concentrate in contiguous intervals rather than isolated single points, while a two‑feature projection indicates that many flags coincide with jointly high values and others arise from atypical value combinations, highlighting multivariate effects.
The study documents practical advantages of Isolation Forest, including minimal distributional assumptions, direct control of alert volume via the contamination parameter, and near‑linear scaling that supports repeated retraining. It also notes limitations, such as sensitivity on small samples due to random tree construction, dependence on threshold calibration that can drift across datasets, and limited inherent explainability of individual alerts. Configuration guidance, robustness checks, and lightweight diagnostics are provided to support deployment and to maintain stable performance under changing conditions.

AI-augmented Scrum: Empirical analysis of the impact of implementing AI tools on the results of IT projects

2025-11-24T17:44:39+02:00

The article presents the results of an empirical study on the impact of integrating artificial intelligence (AI) tools into Scrum project management processes on key performance indicators of IT projects. The influence of AI solutions on sprint duration, development cost, product quality, and stakeholder satisfaction was analyzed. The study is based on a six-month quantitative analysis of 12 IT teams, half of which applied AI assistants (ChatGPT, GitHub Copilot, n8n, Notion AI, etc.) in task management, communication, and analytics. The results indicate that the use of AI assistants in Scrum practices reduces time and financial costs while increasing process transparency and decision quality. In particular, the experimental teams using AI had on average 19% shorter sprints, 14% lower labor costs, and 2.5 times fewer delayed tasks compared to the control teams. Product quality (QA ratings) improved by about 12%, and stakeholder satisfaction increased by 17%. Team managers reported a 35–45% reduction in administrative workload due to automation, and retrospective meetings were ~25% shorter thanks to automated reporting. Utilizing ChatGPT for daily updates and reports enhanced communication transparency with stakeholders. At the same time, certain risks were noted: the need for staff training to effectively use AI tools, the risk of over-reliance on model outputs without critical thinking, and challenges regarding data security and confidentiality when using external AI services. Overall, the findings suggest that AI integration in Scrum can significantly boost team productivity and project outcomes, serving as a catalyst for an AI-Augmented Scrum approach where AI not only supports but also informs management decisions.

Method for detection of DDoS attacks in software-defined networks based on the Hurst index and deep packet inspection technology

2025-11-09T15:34:46+02:00

The article considers the problem of timely detection of DDoS attacks in software-defined networks (SDN), where the centralized controller architecture creates a critical point of failure in conditions of increasing traffic volumes. A combined detection method is proposed, which combines behavioral analysis of traffic using the Hurst index with selective deep packet inspection (DPI). The approach involves dynamic detection of anomalies based on a decrease in the traffic self-similarity index and further refinement of the attack type using signature analysis. The method is integrated into the SDN control plane using CBQ and WRED mechanisms for adaptive queue management. Experimental studies in the Mininet + Floodlight environment confirmed that the combined Hurst–DPI approach provides an increase in attack detection accuracy up to 94%, a reduction in response time by 35%, and a reduction in false positives by 67% compared to traditional methods. The proposed algorithm allows to increase the fault tolerance of SDN networks and maintain the quality of service of critical services in the event of DDoS load.

Analysis of load balancing algorithms in cloud computing

2025-08-23T15:49:39+03:00

This article presents the results of a simulation-based study on the performance of various load balancing algorithms in a cloud service infrastructure using the CloudSim v7.0.0 platform. The simulated environment included one datacenter with 10 hosts, 50 virtual machines, and 200 cloudlet tasks. Seven load balancing algorithms were compared: Round Robin, Throttled Load Balancing, Active Monitoring, Weighted Least Connection, Biased Random Sampling, Min-Min, and Honeybee Foraging. Simulation outcomes were evaluated based on average task completion time, CPU utilization, load variance, and task migration count. Statistical analysis, including ANOVA and pairwise t-tests, revealed significant differences in algorithm efficiency. Based on the findings, further improvement directions for load balancing strategies are proposed, with a focus on adaptive and hybrid approaches.

Alternative approaches for smart contract upgradeability

2025-08-23T13:42:16+03:00

Smart contracts upgradeability is a critical requirement for modern decentralized applications based on blockchain technology, but its implementation remains a technical challenge due to the immutable nature of smart contracts. The Proxy Pattern has become the most widely used solution for implementing upgradability into smart contracts, but it comes with some limitations such as implementation complexity and strict memory layout inheritance. This paper introduces an alternative approach based on dynamic routing with Router Contract, which enables modular upgradeability through centralized address management, offering greater flexibility at the cost of requiring external state persistence. Furthermore, a hybrid architecture is proposed, combining both Proxy Pattern and Router Contract approaches to achieve dual-layer upgradeability – supporting both state-preserving updates and full module replacements. The paper offers a comprehensive evaluation of upgradeability strategies and proposes a versatile solution for evolving smart contract systems.

Technology of quality assessment of critical software systems based on ontological modeling

2025-08-23T13:26:18+03:00

The article investigates the problem of identifying and analyzing the requirements for program systems of critical purpose. For a variety of mission-critical system classes, the task of developing methods for verifying these systems for compliance with the specified requirements within the subject domain and analyzing risks and hazards throughout their life cycle is relevant. Compliance assessment is typically performed by evaluating the overall quality level of flight control software components during the testing phase. To comprehensively analyze the compliance of system attributes with the specified requirements, it is necessary to examine and structure a significant number of objects and functions. To address the problem of systematization within the subject domain, it is proposed to build a formal conceptual model of the system, based on which a set of objects, basic functions, and a quality model to assess the quality level of the system can be technologically formed.

The problem of improving the safety of the operation of objects controlled with automated flight control systems is still relevant. One way to achieve this goal is to ensure the quality and reliable functioning of the flight control systems, which is especially important during the operation of the tolerance control subsystem, which should ensure the complete implementation of control algorithms established for each type of aircraft. In the case of incorrect piloting and operation of units and equipment, there are risks to allow further flights to an unprepared crew or board of the aircraft, which has defects of equipment, which can cause an aviation incident that directly affects the safety of people's lives, as well as the integrity of the aircraft.

The way to solving the problem is the effective use of the ontological model of components of the flight control software systems to create a system of knowledge of the subject area of automated flight control systems. The article analyzes recent research and publications, which shows the effectiveness of using ontology to implement the decision-making process in the context of problematic ambiguity situations. For flight control systems, ontological research apparatus can be effectively used, as the main task of the software is to decide on the receipt of the control object during the flight in the state "within the admission" (regular regime) or "beyond tolerance".

The developed ontological model determines the composition, classification, structure and interconnection of requirements and risks and can forms a technological procedure for creating a quality model of critical PS and the concept of evaluating its characteristics and attributes. Ontological studies and created ontology directly indicate the quality properties that need to be included in the quality models and provide tools for constructing quality models: 1) Target quality model for the developer of the flight control system. 2) Quality in use model for the aircraft operator.

The conducted researches include: 1) Execution of ontological analysis of the subject area of application. 2) Creation of ontology of flight control software, which systematizes and structures knowledge of the concepts and objects of the subject area. 3) Formation of functions of interpretation of ontology, as well as specifications of criteria and restrictions for decision-making problems in critical systems. 4) Formation of specifications of requirements for reliability and safety of flight control systems based on the detection and generalization of components of risk analysis and dangers of functioning of critical software systems. 5)Construction of the target quality model and quality in use model for a given subject area and creation of methodology for evaluating the integral level of quality of flight control software.

The conclusions state that an ontological model and quality models can be used in the life cycle of different classes of control systems (including for certification), since any PS of control are characterized by the presence of a control object, which is controlled by a set of parameters measured by parametric information. The results, obtained in the article, can be apply by developers to create their own quality management system of any critical PS in different subject areas of application.

Further research is proposed to conduct a deeper analysis of the technology of integral assessment of the quality of the PS, especially in terms of normalization of metrics of quality attributes in order to unify them.

A method for expanding the reference model of open systems interaction

2025-08-23T13:21:20+03:00

One type of aviation computer systems formed by a complex of interacting components placed on unmanned aircraft of light and ultralight classes is considered. A method is proposed to reduce the probability of loss of effective, in terms of the cost of production and operation, software and hardware support of the entire system, which can occur with the conscientious use of a multi-level reference model of open systems interaction . The method is based on the use of the morphological analysis method, for which the number of local components is proposed to be equal to the number of levels and sublevels, the search field is expanded for each level autonomously, without taking into account the relationships between levels, and the set of complete solutions of morphological analysis is formed by the classical operation of direct product. The resulting design allows you to determine, for example, by direct search using artificial intelligence tools, those elements (chains) of the display in which each variant of the implementation of the function for the higher levels corresponds to the variants of the implementation of functions and services for the lower levels, which determines the efficiency of the entire system.

Development of a model for forecasting web application vulnerabilities using machine learning methods

2025-08-23T13:16:42+03:00

With the increasing number of cyber threats, risk assessment based on vulnerability data has become a critical task. One of the main approaches involves the use of entropy methods and decision trees, which are applied to evaluate the degree of uncertainty and classify information. This study explores Shannon entropy and the decision tree construction algorithm for analyzing and predicting vulnerabilities. A decision tree model was developed to forecast future vulnerabilities based on a vulnerability dataset. Entropy was also calculated based on threat levels to quantitatively assess the uncertainty in the data.

Weighted checksum based method for fast burst error correction in wireless channels

2025-08-17T12:07:39+03:00

The method for fast correction of burst of errors in wireless data transmission channel for remote control computer systems has been proposed. Method is based on weighted checksums that low computational complexity provided.

The procedures of coding and burst error correction are developed and described in detail and illustrated by numerical example. The comparison analysis of the performance indicators of the proposed method with Reed-Solomon codes has been performed. It has been proved that the developed method provides reducing time of burst errors detection by two orders and time correction reducing by three orders ib comparing to Reed-Solomon codes.

This is of great practical importance in the implementation of control and correcting errors on the terminal microcontrollers of computed remote control systems equipped with single-chip radio modems.

N-phoneme model of speech signal recognition based on hidden Markov processes

2025-08-17T12:02:43+03:00

This paper solves the urgent scientific problem of increasing the probability of recognizing commands and fused speech in radio engineering devices and telecommunications under the influence of distorting factors by developing new recognition models. It is proposed to use hidden Markov processes to conduct a probabilistic description of the one-, three-, and four-phoneme model of speech signal recognition, which makes it possible to theoretically estimate the probability of recognition using each of the models. On the basis of a comparative analysis, the four-phoneme model of speech signal recognition was investigated, which, by improving the three-phoneme model by adding one more state to the model, allows, unlike other models of speech signal recognition, to increase the probability of their recognition. The probability of recognizing speech signals and commands using the four-phoneme method is established, and it is shown that its application in practice with the help of the developed software allows to achieve a probability of 98%. The influence of amplitude and phase distortion of the speech signal on the recognition probability was studied, which showed that the recognition probability decreases when amplitude noise (recognition probability is 81.7%) and phase noise (recognition probability is 92.3%) are introduced into the speech signal. A comparative analysis of the four- and three-phoneme models is carried out, which shows that the recognition probability error of the four-phoneme model is 40% less than that of the three-phoneme model.

Analysis of traffic classification methods in software-defined networks

2025-08-17T11:44:44+03:00

This paper examines traffic classification methods in Software-Defined Networks (SDN). The purpose of this work is to analyze and systematize modern approaches to network traffic classification in the SDN environment.

Traditional classification methods, particularly port-based and Deep Packet Inspection (DPI), are analyzed. Machine learning methods, including SVM, Random Forest, neural networks, and clustering algorithms, along with their advantages and disadvantages, are investigated. The accuracy and efficiency of different classification methods are compared based on published research results.

The authors propose further research on the use of artificial intelligence algorithms in software-defined networks to improve traffic classification accuracy and adaptability to changes in the network environment. The results of comparing different traffic classification methods with a focus on their effectiveness in SDN networks are presented.

Analytical review of methods and technologies for real-time big data processing in IoT infrastructures

2025-08-17T11:37:28+03:00

This paper presents an analytical review of modern methods and technologies for real-time big data processing in Internet of Things (IoT) infrastructures. It explores data stream sources, structural variability, and the key processing requirements such as latency, fault tolerance, and continuous availability. The advantages and limitations of Edge, Fog, and Cloud architectures are discussed, along with Lambda and Kappa approaches for building high-performance IoT systems. Special attention is given to stream processing platforms – Apache Kafka, Flink, Storm, and Spark Streaming – with an evaluation of their scalability, fault tolerance, and ease of deployment. The study highlights state-of-the-art anomaly detection methods in streaming data, including AutoEncoder, LSTM, and Isolation Forest, as well as the use of Complex Event Processing (CEP) for composite event analysis. Real-world applications in smart city systems, industrial automation, and healthcare are examined. The article summarizes current challenges and outlines directions for future research on improving security, adaptability, and efficiency in heterogeneous real-time IoT environments.

Analysis of modern approaches to cybersecurity requirements management during soft-ware implementation

2025-08-16T21:42:36+03:00

This article examines modern approaches to managing cybersecurity requirements during software implementation. It analyzes the regulatory and legal framework, as well as scientific sources governing the requirements for the protection of information systems. Particular attention is paid to the classification of cybersecurity requirements (functional and non-functional), methods of their formulation, documentation, and integration into the software development process. It is determined that cybersecurity requirements management involves not only technical but also organizational and legal aspects, requiring a systematic approach. The role of policies, procedures, technical solutions, and regulations in forming a secure digital environment is considered. The importance of continuous monitoring, updating of requirements, and consideration of the dynamics of cyber threats is substantiated. Furthermore, best practices for managing cybersecurity requirements are presented, along with recommendations for improving the security level of information systems. The obtained results can be used to improve cybersecurity risk management practices, implement security standards, and develop secure software in both the public and private sectors.

Performance enhancement of RGB image convolution using convolution kernel clustering algorithm for ARM64 processor architecture

2025-06-09T10:24:00+03:00

The paper presents a method for improving the performance of RGB image convolution operation on the ARM64 platform using a convolution kernel element clustering algorithm. The proposed approach is based on vectorization of computations using NEON64 SIMD instructions and grouping of non-zero kernel elements with the same sign for efficient skipping of operations with zero elements. A mathematical model of vectorized convolution operation has been developed, which takes into account the specifics of sparse convolution kernel matrices. Experimental study on the Orange Pi 5 Pro platform demonstrated significant acceleration compared to the cv::filter2D() function of the OpenCV library: for medium-sized kernels (7×7 – 11×11), an acceleration of 5.0–9.7 times was achieved, for large kernels (12×12 – 15×15) – 1.7–5.5 times. The proposed method is particularly effective for processing high-resolution images and can be applied in real-time systems on single-board computers with limited computational resources.

Computerized planar tomography system for diagnostics of combat injuries

2025-06-08T23:21:12+03:00

In this work, the use of mobile digital X-ray tomosynthesis complexes for diagnosis of shrapnel wounds in front-line hospitals was substantiated. During the research, an increase in the contrast of pathological images by 6...8 times was experimentally obtained in mobile X-ray complexes with tomographic information technology compared to X-ray images. A modeling stand was created in which a working sample of a veterinary mobile device with tomosynthesis was used and a measuring phantom was developed that simulates shrapnel lesions of the extremities. This made it possible to obtain reliable measurement results.

Package coding into ad-hoc networks

2025-06-08T23:15:16+03:00

Based on the analysis of known approaches to coding network traffic in wireless data networks, an attempt has been made to substantiate the possibility and feasibility of using network coding in special-purpose networks (Ad-hoc networks). At the same time, the features of the implementation of some coding algorithms were considered from the point of view of the limitations caused by the environment of lossy networks. It has been shown that the implementation of packet coding methods in lossy networks, which are essentially Ad-hoc networks, is not only possible, but also useful in terms of increasing the reliability and efficiency of data transmission, but the answer to the question of how these methods work in practice requires additional implementation.

Security mechanisms for AODV and E-AODV protocols against black hole and gray tunnel attacks

2025-06-08T23:00:57+03:00

This paper investigates the security vulnerabilities of the AODV (Ad hoc On-Demand Distance Vector) and E-AODV (Enhanced AODV) routing protocols in mobile ad hoc networks (MANETs), focusing on two common attack types: black hole and gray tunnel attacks. A comprehensive analysis is conducted to examine how these attacks affect routing reliability, packet delivery, and network performance. E-AODV is evaluated as an improved protocol incorporating adaptive trust mechanisms, response timing analysis, and behavioral monitoring to mitigate malicious node interference. A simulation framework is implemented using a Python-based environment, where various network sizes (10–100 nodes) are tested. Key performance indicators such as Packet Delivery Ratio (PDR), End-to-End Delay, and Packet Loss Rate (PLR) are used to compare protocol behavior. Results show that E-AODV significantly enhances resilience to attack-induced disruption, outperforming standard AODV in both delivery efficiency and security reliability across diverse MANET topologies.

Problems and requirements when creating an automated heating system for residential premises with increased energy efficiency

2025-06-08T22:53:54+03:00

The article discusses in detail the problems and requirements for creating an automated residential heating system with increased energy efficiency and explores the transformational potential of such automated systems.

The discussion begins with an overview of the importance of residential heating systems, emphasizing the need for increased energy efficiency in the face of environmental concerns and rising energy costs. The introduction of automated systems is introduced as a promising solution, leveraging advanced technologies such as smart sensors, machine learning algorithms, and IoT connectivity.

The article delves into the challenges posed by traditional heating methods, including common inefficiencies and environmental impacts, paving the way for a comprehensive understanding of the advantages offered by automated systems. Additionally, the article emphasizes the role of government incentives and rebates in promoting energy-efficient home upgrades, providing practical tips for homeowners considering the adoption of automated heating systems. The exploration of emerging technologies, advancements in energy efficiency, and the integration of artificial intelligence offers a glimpse into the future of residential heating.

In conclusion, the article underscores the benefits of automated heating systems, encourages homeowners to embrace energy-efficient solutions, and provides insights into the evolving landscape of residential heating systems. The adoption of these technologies is presented not only as a means of improving personal comfort and reducing energy bills but also as a critical step towards building a more sustainable and environmentally conscious future.

Modular multiplication on invariable number with group processing of multiplier bits and Montgomery reduction

2025-06-08T22:45:28+03:00

Developed and studied a method for accelerating the computational implementation of modular multiplication by a constant number as a component of modular exponentiation from high bits – the basic computational operation of asymmetric cryptography. The acceleration is achieved by using precomputations to combine group processing of k bits of the multiplier and the corresponding Montgomery reduction. Numerical examples are given to illustrate the work of the proposed method.

It has been theoretically proved and experimentally confirmed that the developed method allows accelerating modular multiplication by a constant number by a factor of k in comparison with known methods.

Method for reconfiguring a distributed information system after partial physical destruction

2025-06-08T19:08:10+03:00

The article addresses the problem of restoring the functionality of distributed information systems (DIS) that have suffered partial physical destruction and proposes a method for their reconfiguration. In cases of physical damage caused by external or internal factors, ensuring the continued operation of such systems is of critical importance. The loss of connectivity between elements can lead to decreased computational efficiency, disruption of data exchange, and partial or complete loss of system control.

The proposed method aims to restore connectivity and adapt the DIS to new operating conditions by reconfiguring its structure. The core idea of the approach is to identify alternative communication routes between system elements to compensate for lost connections. The method is based on modeling the system structure as an undirected graph, where damaged connections are represented as missing edges, and functionality is restored by constructing a set of finite simple chains.

The relevance of this research is driven by the increasing dependence on distributed computing systems in critical sectors such as military, energy, transportation, and telecommunications. Efficient recovery of their functionality minimizes the impact of destruction, enhances resilience, and ensures operational continuity even in crisis situations. The proposed method is universal and can be adapted to a wide range of distributed systems, including both computer networks and other complex cyber-physical systems.

The scientific novelty of the method lies in its ability to perform a comprehensive search for all possible connectivity restoration paths while minimizing computational costs. This is achieved by pre-narrowing the solution space, allowing the method to efficiently scale for systems with a large number of nodes. An additional advantage is its applicability for modeling system fault tolerance scenarios and predicting the consequences of destruction, making it valuable for cybersecurity strategy development and emergency response planning.

Intelligent agents in the fight against global threats

2025-06-08T19:01:00+03:00

This paper explores the application of intelligent agents in combating tuberculosis, which remains a significant public health threat. A cellular automata-based disease spread model is proposed, enabling the analysis of infection dynamics while considering socio-economic and demographic factors. Simulations conducted for Lviv and Stockholm reveal significant differences in the speed of disease transmission depending on the initial level of latent infection.

In Lviv, the projected number of active tuberculosis cases reaches 308 per 100,000 people within ten years, highlighting the need for effective control measures. In Stockholm, where the initial infection rate is much lower, the disease spreads gradually, driven by migration processes. Model validation using official WHO data confirmed its accuracy, allowing it to be used for epidemic forecasting and the development of disease control strategies.

The results emphasize the importance of early diagnosis, active case monitoring, and the consideration of migration processes in shaping preventive measures. The use of intelligent agents can significantly improve epidemic management and enhance the effectiveness of public health strategies.

Method of distributed calculation on multi-core processors exponential on Galois fields for cryptographic applications

2025-06-08T18:18:52+03:00

The organization of rapid parallel calculation of the exponential on Galois fields for cryptographic applications on multi-core processors is proposed.

The proposed approach to parallelization is based on the possibility of exponentiation a number to any degree of two on Galois fields by using precomputations over the same time. Two schemes of organization of parallel computing processes of exponentiation on Galois fields of were proposed. For both schemes, the procedures for calculating partial exponential on k processor cores are developed in detail and numerical examples that illustrate their work are given.

It has been theoretically proven and experimentally confirmed that proposed method of distributed exponentiation on Galoise fields on multi-core processors allows, due to the parallelization and use of precomputation, to reduce the calculation time by 2×k times.

Algorithm for orchestration of encrypted traffic in SDN networks

2025-06-08T18:13:06+03:00

The article addresses the issue of overload in Software-Defined Networks (SDN) under increasing volumes of encrypted traffic, which complicates its classification and reduces the efficiency of traditional load balancing methods. The proposed algorithm performs preliminary traffic classification using DPI, NBAR, and Cisco AVC technologies, followed by the application of restriction policies for non-priority flows. The solution is based on the M/M/1/K queueing model and considers the characteristics of priority traffic. The proposed approach improves network fault tolerance, enhances the quality of service for critical applications, and reduces data loss in high-load environments.

Method of traffic orchestration in software defined networks considering quality of service parameters

2025-06-08T16:53:33+03:00

This study is dedicated to the method of traffic load balancing in computer networks using SDN technology. The path selection criterion is proposed that enables more balanced loading of data transmission channels while ensuring compliance with specified QoS parameters.

The modified method of traffic orchestration is proposed to reduce the time required to form a set of partially overlapping paths and to simplify the traffic balancing procedure. The availability of this path set enables the replacement of the rerouting process with a path reconfiguration procedure during data transmission.

The problem of information protection in electronic document management systems of medical organizations

2025-06-08T16:36:16+03:00

The article is devoted to the study and processing of methods for creating electronic document management systems of medical organizations, increasing the level of protection of documents and information stored and processed in this system. For this purpose, International and local experience in the field of legislation on the protection of documents containing personal and medical information of patients in medical organizations was commented in the article, the issue of creating a distributed structured electronic document management system for medical organizations was considered, various protection methods were used to ensure the confidentiality of personal and medical information of patients, it was proposed to use the authentication procedure and secure encryption methods.

Performance analysis of parking occupancy prediction models in an IoT-based intelligent system

2025-06-08T16:13:35+03:00

The article investigates the effectiveness of parking occupancy forecasting models within an Internet of Things (IoT) intelligent system operating on cloud computing principles. The relevance of parking space management in modern urbanization is considered. The architecture of an analytical system is presented, which includes IoT sensors for data collection, a cloud platform for data processing and storage, and a forecasting module. Three time-series forecasting models are described in detail: LSTM (Long Short-Term Memory), Prophet, and SARIMA (Seasonal Autoregressive Integrated Moving Average). An experimental study of these models was conducted on a parking occupancy dataset, including their training, testing, and evaluation using MAE, RMSE, and MAPE metrics. A comparative analysis of the accuracy of the models was performed, identifying the advantages and disadvantages of each in the context of forecasting parking zone occupancy. The research results demonstrate the potential of using machine learning models to optimize urban parking infrastructure management and can be used for further improvement of intelligent transport systems.

Use of complementary filter for identification of motion parameters of elements of spherical parallel mechanism

2025-06-08T16:05:15+03:00

The article proposes to use a complementary filter in the SPM control system for processing data from the accelerometer and gyroscope. A complementary filter combines data from an accelerometer that measures linear acceleration, which can be used to determine the angle of inclination in a static position, and a gyroscope, which measures angular velocity, which is used to integrate and determine the angle of rotation. The filter calculates a weighted average of these two measurements, with the weights chosen to minimize the effects of noise and drift from each sensor.

An algorithm for calculating the angle of inclination of the SPM platform was developed, where the final value of the angle of inclination is the sum of the integrated value of the gyroscope and the instantaneous value of the accelerometer. At each stage of integration, the integral of the angle of inclination is corrected using the readings of the accelerometer. An experimental verification of the effectiveness of the complementary filter was performed to identify the motion parameters of the elements of the spherical parallel mechanism.

Analytical review of actuators and adaptive control in bionic upper limb prostheses

2025-06-08T13:36:35+03:00

The article considers modern types of actuators used in bionic upper limb prostheses, with an emphasis on their classification, technical characteristics, advantages and limitations. A comparative analysis of pneumatic, hydraulic, electric, as well as innovative electromechanical actuators based on smart materials is carried out. Special attention is paid to actuator control systems: from classical methods, in particular PID controllers, to modern approaches using neural networks, machine learning and adaptive algorithms. The human-prosthesis interfaces and the challenges associated with noise in biological signals and the need for individual adaptation of the control system to the user are considered. The material is of practical importance for engineers and researchers involved in the development of a new generation bionic prostheses.

Hybrid models for detecting shadowburst anomaly in industrial IoT traffic

2025-06-07T22:16:35+03:00

The increasing deployment of Industrial Internet of Things (IIoT) systems within Industry 4.0 environments has introduced new cyber-physical vulnerabilities, particularly in the form of stealthy and short-lived anomalies that evade traditional detection mechanisms. This paper introduces and formalizes a novel anomaly type, referred to as ShadowBurst, which consists of protocol-conformant, high-frequency microbursts embedded in otherwise stable traffic streams. We propose a hybrid detection architecture that integrates Kalman filtering for temporal state estimation with machine learning techniques, specifically Isolation Forest, for residual-based outlier detection. The detection function is further enhanced by incorporating statistical scoring and behavioral profiling to improve anomaly visibility. Simulation results confirm that this hybrid Kalman–ML approach enables effective identification of ShadowBurst anomalies in time-sensitive IIoT traffic, addressing gaps left by signature-based and purely statistical models. The proposed model demonstrates high responsiveness to low-duration, protocol-mimicking threats and supports real-time deployment in smart manufacturing environments.

An approach to optimizing the microservices deployment model in a highly loaded environment

2025-03-17T00:38:42+02:00

The article proposes the use of a LB-Diversity algorithm for deploying microservices to increase system stability and optimize resources in cloud architectures. Traditional microservice systems often face node failure issues and security vulnerabilities due to homogeneous deployment of identical service instances. The proposed approach addresses these issues by combining load balancing with deployment diversity, where different versions of microservices are distributed across multiple nodes. This strategy reduces the risks associated with a single point of failure and mitigates the impact of attacks targeting common vulnerabilities in containerized environments.

To analyze the experiments, our own coefficients for assessing load balancing and system fault tolerance were proposed, which confirmed the results of previous studies on the choice of the LB-Diversity algorithm compared to traditional deployment methods. The diversity of deployment approach effectively limits the damage from potential security breaches by ensuring that only a small portion of the system is affected. This model offers a practical solution for enterprises that require reliable microservice architectures with high availability and security in dynamic environments with limited resources.

A method for fast implementation of Fourier transform with secured use of cloud computing

2025-03-15T00:08:26+02:00

The article proposes a method for fast implementation of the discrete Fourier trans-form (DFT) on a terminal microcontroller (TMC), which is based on additive masking of signals and is distinguished by the fact that it updates the mask by changing its bit layers, which allows to update the mask before processing each input signal. The basic procedures of the method are theoretically justified and described in detail. An analysis of the effec-tiveness of the proposed method of homomorphic encryption of signal readings from a real object was performed. It is shown that due to the use of a faster procedure for updating the mask by changing its bit layers, the level of protection of signal readings against attempts to illegally restore them in the cloud is significantly increased. This is achieved by changing the mask for each signal.

It is also shown that the proposed method can reduce the time of encryption and de-cryption by two times. The conducted experimental studies generally confirmed the results obtained by theoretical evaluation of efficiency.

Method for synthesis of fault-tolerant topologies with implicit clusters based on de Bruijn transformations in redundant numeral systems

2025-03-15T00:03:46+02:00

The work is devoted to the development of a method for synthesizing topologies based on de Bruijn transformations in redundant numeral systems, which allows synthesizing fault-tolerant topologies of a given order, including those with implicit clusters. A method for forming such clusters and a method for studying the survivability of topologies using dynamic determination of failure probability based on betweenness centrality are also developed.

The proposed comprehensive approach allows us to synthesize graphs that, on the one hand, contain less redundancy, and on the other hand, have higher survivability due to better use of the available redundancy, which allows us to increase fault tolerance with lower costs and ensure better efficiency.

Mathematical model of measuring details on a coordinate measuring machine

2025-03-14T23:52:10+02:00

The paper explores the significance of measurement accuracy in coordinate measuring machines (CMM), highlighting their role in modern quality control processes. Key sources of measurement errors, including geometric, thermal, kinematic, and random factors, are analyzed. The study presents a novel approach for developing error correction algorithms aimed at enhancing measurement accuracy. The mathematical model developed considers the primary error sources and provides a structured framework for their correction. An algorithm for error correction is proposed, detailing the integration of systematic error adjustments into the measurement process. The algorithm is designed for real-time application, offering dynamic adaptability to varying operational conditions.

The implementation of the proposed algorithm in CMM software demonstrates significant improvements in measurement precision. Results of modeling validate its effectiveness across diverse conditions. The paper concludes by emphasizing the scientific contribution of the proposed methodology, which combines theoretical advancements with practical applications to enhance the accuracy of CMM measurements.

ReliefF functional selection for traffic anomalies detection

2025-03-13T00:35:04+02:00

This paper addresses the problem of network traffic anomaly detection using machine learning techniques and feature selection optimization. As modern computer networks become increasingly complex and data volumes grow, efficient threat detection requires fast and accurate traffic analysis algorithms. A major challenge lies in distinguishing normal network activity from potentially malicious attacks, which may be disguised as legitimate traffic.

The study explores the use of the ReliefF method for selecting key network traffic features and its impact on the performance of the Random Forest classification model. The focus is on optimizing feature selection to reduce computational costs and speed up training while preserving essential indicators of anomalous activity. Special attention is given to evaluating system performance under different feature selection scenarios, allowing for an optimal trade-off between accuracy and processing speed.

The proposed approach is designed for automated intrusion detection systems (IDS) operating in real-time, requiring rapid responses to potential threats such as DDoS attacks, port scanning, unauthorized access, and exploits. The research aims to enhance the efficiency of existing network traffic analysis methods and adapt feature selection algorithms for high-load network environments.

Routing method in wireless IoT networks with high device density

2025-03-12T03:28:17+02:00

This research is based on the assumption that in the immediate vicinity of the calculated coordinates are located transit nodes, which are located along a straight line connecting the initial and final nodes of the route. This assumption is valid for networks with a high density of nodes. For networks with a limited number of nodes, it is advisable to focus on solving the problem of determining the required number of nodes of the route based on the distances between them.

The proposed method reduces the use of network resources by selecting a near-optimal route, taking into account the influence of the transmission speed on the level of the useful signal power and interference from neighboring nodes on the route. For example, it allows you to increase the overall data transfer rate compared to the approach based on determining the shortest route based on the number of transit sections.

Research on the modernization of the data transmission standard ARINC629

2025-03-12T03:08:06+02:00

As of 2024, due to its high reliability and efficient data transmission between various aircraft systems, enhancing flight safety and efficiency, the ARINC 629 digital data bus exchange standard is still used in various aircraft models: Boeing 777-200/300/ER/F, Boeing 787-8/9, Boeing 747-8, Airbus A350-900/1000, Airbus A380-800, Embraer E-190/E-195, Bombardier CRJ900/1000, and Sukhoi Superjet 100. The aircraft listed above utilize ARINC 629 for data transmission among different systems: Flight Management System (FMS), Engine Control System (ECS), Air Conditioning (AC), Lighting (L), and Communication (C). The article suggests replacing the signal transmission material with fiber optic technology. Indeed, there are upgrades to this standard in the form of ARINC 818 [11, 12]. However, the implementation of these standards tends to be more specialized and less flexible.

Improvement of the mathematical model of the routing table taking into account the psychophysiological mechanisms of the reaction of the computer network administrator

2025-03-12T02:58:46+02:00

The presented work considers the problem of optimizing the topological structure of the network. In this case, an improved logarithmic function of the administrative distance was used, which corresponds to the fundamental law of the analysis of “man-machine” systems – the Weber-Fechner law. The results of calculations and computer modeling indicate that the intuitive reactions of the computer network administrator to changes in the logarithm of the administrative distance are accelerated when applying the administrative distances of the Weber-Fechner law, an approximately 10...15 percent acceleration of the intuitive reactions of the network administrator to sudden excitations is observed, which gives an additional gain in reducing the delays of adaptation to route changes by approximately 2%-3%.