EUVIMEDEUVIMED entdecken
Crossref

Markov model for forecasting the development of cyberattacks in computer networks

T.O. Vasylenko, D.Yu. Gorelov, O.O. Ivanova

Radiotekhnika · 25.07.2026 · (225) · 99-103

DOI: 10.30837/rt.2026.2.225.09 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Radiotekhnika, online erschienen am 25.07.2026. Verfasst von T.O. Vasylenko und 2 weiteren Autor:innen.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Modern computer networks are characterized by a high level of complexity and a continuous growth in the number of cyber threats, which necessitates the development of effective methods for detecting and predicting the progression of cyberattacks. Many modern cyber incidents are multi-stage in nature and are carried out as a sequence of interrelated attacker actions, ranging from reconnaissance and network scanning to vulnerability exploitation, persistence establishment, and data exfiltration. Traditional intrusion detection approaches based on signature or rule-based methods typically respond only to known attack patterns and do not provide the capability to predict the further development of an incident. Therefore, the application of probabilistic models capable of describing the dynamics of transitions between system states under conditions of uncertainty and partial information is becoming increasingly relevant. One promising approach to modeling the development of cyber threats is the use of Markov processes, which allow the sequence of attack stages to be formalized as a stochastic process with certain probabilities of transitions between states. This approach makes it possible not only to analyze already observed security events but also to estimate the probability of further attack progression, thereby providing a basis for timely response and improving the effectiveness of cybersecurity systems. The use of Markov models in combination with modern attacker behavior classifications, particularly the MITRE ATT&CK framework, enables a more accurate description of the stages of complex attacks in computer networks. This paper proposes a Markov model for predicting the progression of cyberattacks, in which the system operation process is represented as a discrete sequence of states corresponding to different stages of an attack or normal system operation. Based on the analysis of security event logs, a transition probability matrix between states is constructed and used to estimate the most probable next stage of an attack as well as to forecast its further development several steps ahead. The experimental evaluation of the model was performed using public network traffic datasets containing labeled attack scenarios. The results of the study show that the proposed model can predict subsequent stages of attack development with sufficiently high accuracy and can be used as a decision-support tool in security monitoring systems. The application of the Markov approach makes it possible to assess the risk of system transition to critical states and to implement preventive protection measures for computer networks in a timely manner.

Zitieren

APA

Vasylenko, T., Gorelov, D., & Ivanova, O. (2026). Markov model for forecasting the development of cyberattacks in computer networks. Radiotekhnika, 99-103. https://doi.org/10.30837/rt.2026.2.225.09

Vancouver

Vasylenko T, Gorelov D, Ivanova O. Markov model for forecasting the development of cyberattacks in computer networks. Radiotekhnika. 2026:99-103. doi:10.30837/rt.2026.2.225.09

BibTeX
@article{Vasylenko2026,
  author = {Vasylenko, T.O. and Gorelov, D.Yu. and Ivanova, O.O.},
  title = {Markov model for forecasting the development of cyberattacks in computer networks},
  journal = {Radiotekhnika},
  year = {2026},
  number = {225},
  pages = {99--103},
  doi = {10.30837/rt.2026.2.225.09}
}
Kostenlos starten →