Кафедра інформаційно-мережної інженерії (ІМІ)
Постійний URI для цієї колекції
Перегляд
Останні подання
Публікація Дослідження методів ідентифікації людини за її зображенням на основі використання нейромереж(ХНУРЕ, 2025) Сливенко, Є. І.The work aims to evaluate the effectiveness of methods for identifying a person from his or her image using neural networks. The main approaches to analyzing video streams and methods of automatic object recognition based on deep learning algorithms are considered. Studies have shown that the use of neural network algorithms can increase the accuracy of fire detection by 50% compared to traditional methods.Публікація Використання методів багатокритеріальної оптимізації для вибору телекомунікаційного обладнання(ХНУРЕ, 2025) Скорик, Ю. В.; Ліннік, М. В.The paper considers theoretical and practical aspects of the application of multi-criteria optimization methodology to select design solutions that are optimal according to a set of dependent and contradictory quality indicators, in particular, for L2 level managed switches taking into account 5 quality indicators: price; number of ports; number of combined ports; buffer memory capacity, throughput; number of forwarded packets. At the same time, a Pareto subset of 22 valid variants is first selected, which included 9 variants. The remaining variants of the system are unconditionally the worst and are excluded from further consideration. Then, the Pareto subset is narrowed down to a single system variant by methods based on the theory of fuzzy sets, utility theory, and lexographic method.Публікація Аналіз видів корпоративних мереж(ХНУРЕ, 2025) Скорик, Ю. В.; Плех, О. А.This work analyzes the types of corporate networks, namely departmental networks and campus networks. The advantages and disadvantages of corporate networks are considered. Departmental networks effectively connect and coordinate the work of individual departments, while campus networks integrate larger geographically dispersed structures, creating a single information system for the company. Both types of networks increase productivity, help optimize business processes, and provide reliable communication across the enterprise.Публікація Огляд можливостей застосування засобів штучного інтелекту щодо процедур тестування програмного забезпечення інфокомунікаційних систем(ХНУРЕ, 2025) Прохоров, О. І.; Бондар, Д. В.Intensive development of modern infocommunication systems requires high quality testing to ensure efficiency, security and productivity. The main types of testing are functional, load, security testing, globalization and automated testing. The use of artificial intelligence significantly improves the quality of testing, reduces costs and speeds up processes. Currently, there are AI tools: ChatGPT, Gemini, Mistral and Claude3, which demonstrate quite powerful capabilities in log analysis, test scenario creation and test automation. The most effective for testing infocommunication systems is ChatGPT, which provides the best support for test scenario generation and process automation.Публікація Особливості проектування корпоративної мережі підприємства(ХНУРЕ, 2025) Скорик, Ю. В.; Бєлих, К. В.This work considers the principles of designing corporate networks of an enterprise. It has been revealed that when creating a corporate network for a small manufacturing company, the most available, undemanding hardware modeling tools are used. Сreating an architecture for large companies with a wide range of activities implies the use of powerful tools. When planning an enterprise corporate network, as many of the requirements specified in the standards as possible should be taken into account for the best long-term functioning of the network.Публікація Порівняння засобів телекомунікацій за допомогою методу аналізу ієрархії(ХНУРЕ, 2025) Скорик, Ю. В.; Кобеляцький, В. В.This work considers the theoretical and practical peculiarities of applying the method of hierarchy analysis to select the optimal tablet from some set of options when choosing the preferred option of telecommunication means, taking into account a set of quality indicators. As a result of judgment processing, estimates of the corresponding eigenvectors and priority vectors for different levels of the comparison hierarchy were calculated. These estimates were used to calculate the values of the global priority vector components. According to the method of hierarchy analysis according to the maximum value of the global priority vector component, the preferred tablet was selected taking into account the quality indicators.Публікація Порівняльний аналіз засобів виявлення шкідливих мережних пакетів(ХНУРЕ, 2025) Матвейченко, А. Є.This work presents a comparative analysis of current network packet analyzers, addressing the increasing complexity of network infrastructures and the growing demand for effective cybersecurity monitoring. In light of rising cyber threats, the research emphasizes the necessity for a comprehensive approach to traffic analysis that not only deciphers detailed protocol data but also offers intuitive monitoring for a broad range of users. An evaluation of popular tools, such as Wireshark, Tcpdump, Nmap, Kismet, NetworkMiner, and SolarWinds Network Performance Monitor, reveals unique strengths and limitations pertinent to various operational environments. The work identifies key criteria for efficiency, including performance, analysis depth, user-friendliness, protocol support, cost, and platform independence.Публікація Безпека бази даних у високо регульованих галузях(ХНУРЕ, 2025) Ніколаєв, А. О.; Кривенко, С. А.The document discusses the importance of database security in highly regulated industries, particularly focusing on the use of Amazon RDS and various methods to enhance data protection. Users must evaluate data confidentiality and ownership standards in compliance with current laws, especially in healthcare and finance sectors. Cloud technologies like AMAZON RDS provide new research opportunities. Amazon RDS is suitable for OLTP, efficiently handling large volumes of transaction data, such as banking transactions, by securely storing and updating transaction details. A variety of methods are recommended to manage access and enhance security for Amazon RDS, including using virtual private clouds, SSL/TLS connections, encryption, and database security features.Публікація Метод визначення степенів утворюючих поліномів псевдовипадкових двійкових послідовностей(ХНУРЕ, 2025) Олещук, Д. В.; Бондар, Д. В.This work is devoted to the description of the method of programmatic determination of the powers of the generating polynomials of pseudo-random binary sequences, according to which the cell numbers of the shift register of the sequence generator are selected. The algorithms for creating tuples of register bit numbers and selecting from them those that can be used to generate pseudo random binary sequences are considered. There is a link to the developed software module. The advantages and disadvantages of the considered method are formulated.Публікація Аналіз використання технологій блокчейн для розподіленої обробки і зберігання даних(ХНУРЕ, 2025) Кураков, Я. С.; Колтун, Ю. М.This work analyzes the use of blockchain technologies for distributed data processing and storage on the example of Ethereum and Hyperledger platforms. In the course of the analysis, the main problems associated with these processes, as well as with the transmission of sensitive data through data transmission networks, are highlighted. It is shown that most of them can be mitigated by applying appropriate algorithms, such as the Proof-of-Work (PoW) algorithm. The Network Simulator (NS-3) together with the Bitcoin Simulator framework has been used to model the environment. PoW blockchains containing several hundred network devices in close geographical proximity have been found to achieve the highest performance.Публікація Вибір оптимального варіанту багатофакторної біометричної автентифікації в інфокомунікаційних мережах(ХНУРЕ, 2025) Леонова, Ю. А.; Чеботарьова, Д. В.The paper examines various types of biometric authentication and their features, in particular various methods of static and dynamic biometrics. The current state, development trends and promising areas of application of biometric technologies are analyzed. The most common biometric authentication methods are considered: by face geometry, fingerprint, voice, iris and venous pattern. For these biometric methods, an assessment of quality indicators was performed and a multi-criteria selection of the optimal method was carried out using a set of quality indicators. The optimal method was the method of biometric authentication by venous pattern. Some options for multi-biometric methods were also considered. The optimal method was a method that uses a combination of the following biometric features for user authentication: fingerprint and venous pattern.Публікація Методи виявлення у системах виявлення вторгнень(ХНУРЕ, 2025) Лійко, Д. О.; Іваненко, С. А.This work focuses on intrusion detection systems (IDS) used to identify malicious activities in computer systems and networks. The main types of IDS, namely, network-based (NIDS) and host-based (HIDS), are considered. The key detection methods, including signature-based and anomaly-based approaches, are analyzed. The advantages and limitations of each method are discussed, with particular attention to the adaptability of anomaly detection in identifying new and unknown threats. The role of IDS in modern cybersecurity infrastructure is highlighted, demonstrating its importance in threat detection and mitigation.Публікація Причинно-наслідковий штучний інтелект і інтернет речей(ХНУРЕ, 2025) Кирсанов, О. О.; Кривенко, С. А.The document discusses an experiment involving dynamic administration of medications using causal artificial intelligence, specifically Amazon SageMaker, to optimize treatment protocols. The study aims to experiment with dynamic administration of medical drugs to patients using causal artificial intelligence to optimize treatment protocols and improve overall treatment outcomes. The experiment utilized Amazon SageMaker methods, which provide tools to automate processes, optimize workflows, and create machine learning models. The results indicated that the proposed AI model could optimize medication dosages, leading to a significant improvement in the overall treatment effect by more than 20%Публікація Принципи побудови обчислювальних мереж(ХНУРЕ, 2025) Скорик, Ю. В.; Козін, А. О.This work is devoted to the analysis of the construction of local and corporate communication networks. The main schemes of the department, campus and megalopolis networks are considered. A universal cable system is constructed in which horizontal and vertical subsystems are present. The purpose of the Data Link layer and its sublayers is analyzed in detail. The components of the subsystems of the structured cable system are considered.Публікація Створення та впровадження фреймворку для автоматичного тестування інфокомунікаційних додатків(ХНУРЕ, 2025) Красніков, В. О.; Чеботарьова, Д. В.This discusses the development and implementation of an automated testing framework for infocommunication applications. Automated testing significantly reduces testing time, enhances software stability, and minimizes human-related risks. The proposed framework integrates seamlessly into the development process, ensuring systematic and repeatable test execution. The study highlights the advantages of automation, including CI/CD integration, scalability, and efficient resource utilization. Additionally, it emphasizes the importance of selecting appropriate tools, such as Selenium or Appium, based on project requirements. The research underscores the strategic role of test automation in improving software quality and accelerating product release cyclesПублікація Порівняльний аналіз технологій для організації мережі доступу підприємства критичної інфраструктури(ХНУРЕ, 2025) Кузьмін, А. А.; Колтун, Ю. М.This work is devoted to the comparative analysis of technologies for organizing access networks in critical infrastructure enterprises. Various approaches, including wired and wireless solutions, were considered. The study examines the advantages and disadvantages of Ethernet, fiber-optic networks, Wi Fi, LTE, and 5G technologies. A comparative assessment of security, reliability, and deployment costs was conducted. It has been found that hybrid solutions combining wired and wireless networks provide the most balanced approach, ensuring both high performance and flexibility while maintaining the necessary level of security.Публікація Аналіз принципів створення API у хмарній інфаструктурі(ХНУРЕ, 2025) Бондар, Н. О.; Колтун, Ю. М.A key solution for ensuring seamless integration and interaction in cloud infrastructure is the API. APIs provide a standardized set of protocols and tools for developing, integrating, and managing software applications, enabling efficient communication between distributed services. This paper analyzes the fundamental principles of API development in cloud ecosystems, focusing on best practices for creating scalable, secure, and reliable services, and explores the advantages of API-driven architectures, including cost optimization, enhanced interoperability, and reduced development complexity. A case study of AWS API Gateway illustrates how managed API services facilitate the design, deployment, and security of cloud-based applications.Публікація Аналіз потенційних ризиків комплексної системи протидії інформаційним загрозам(ХНУРЕ, 2025) Даниленко, М. С.; Соколюк, В. А.; Клушин, Д. Д.; Безрук, В. М.The modern technological base provides a fairly wide range of software and hardware tools to counteract malicious influences that pose a potential threat to network infrastructure and data in file storage. However, based on the existing world experience in protecting against cyber threats, it can be stated that only a multi-level complex system can provide a significant level of security. On the other hand, any complex system has potential vulnerabilities that can be used by an attacker to successfully implement an attack. Therefore, to maintain network information security, it is necessary not only to deploy a protection system, but also to audit its possible vulnerabilities and localize each of themПублікація Розпізнавання сигналів за допомогою нейронних мереж(ХНУРЕ, 2025) Дубограй, Д. Д.; Колтун, Ю. М.Human sensory organs function as multi-component receptors that perceive images and generate multi-channel pulse signals, and recognizing them is crucial for the development of next-generation neural networks known as neuromorphic intelligent systems. Spiking Neural Networks (SNNs) are more effective because they can detect dynamic signals without converting them into feature vectors. This work presents a method for detecting multi-dimensional signals using SNNs, detailing their structure, models and training procedures. SNNs can be trained digitally or by special circuits that adjust synaptic weights without digital processorsПублікація Розробка інтернет-магазину streetline(ХНУРЕ, 2025) Гордієнко, Є. С.The correct structure of an online store is a guide for customers. Without it, visitors can get “lost” and go to another site. The competition in the online market is enormous, and every little thing can be critical for business. The site structure should be worked out at the stage of designing the store. Before launching the portal, it is recommended to conduct tests, for example, with the help of a focus group, to assess the site's usability. Developing the structure of an online store is a fairly simple task if it is handled by an experienced specialist. You shouldn't save effort, time, and money at this stage to avoid high costs in the future. Translated with DeepL.com (free version).