Кафедра комп’ютерної інженерії та охоронних систем

Постійне посилання на фондhttps://repository.lntu.edu.ua/handle/123456789/76

Переглянути

Результати пошуку

Зараз показуємо 1 - 2 з 2
  • Item type:Наукова стаття,
    Socket performance influence on data processing intensity in a virtual machine cluster with heterogeneous conditions
    (Lutsk: LNTU, 2024-06-21) Melnyk, Vasyl; Bahniuk, Nataliia; Roiko, Oleksandr; Bortnyk, Kateryna; Kizym, Svitlana
    The paper presents the dependences of data processing in heterogeneous conditions taking into account the impact of socket performance. As high-performance, sockets are proposed on the virtual interface architecture to use a component basis in order to support the application performance for intensive data processing. High-performance applications, using the traditional Linux-based TCP/IP interface, require guarantees of performance and its scalability to adapt for heterogeneous networks. In case of using highly effective protocols for their operation, some additional methods should be used at the user level, including for sockets on the virtual interface architecture. The limitation on the highperformance socket layers is tyed with the application developing completion to full performance salvation during the TCP/IP data transfer process. There also revealed that small-block load balancing has been found to increase heterogeneous network adaptation.
  • Item type:Наукова стаття,
    Development of network traffic monitoring system elements using Deep Learning
    (Piscataway, NJ: IEEE, 2024) Melnyk, Vasyl; Bahniuk, Nataliia; Bortnyk, Kateryna; Kondius, Inna; Zubovetska, Nataliia; Kondius, Kostiantyn
    In this article we develop a network traffic monitoring system elements using deep learning that demonstrates the high effectiveness of deep learning in detecting and analyzing network traffic, which contributes to ensuring the security and stability of corporate networks. In this context we develop algorithms for detecting and countering potential threats in network traffic, providing deeper analysis and effective response to cyber threats