Stability and Sensitivity Analysis of Cyberattack Propagation Models in Computer Networks

Abstract

This study develops a compartmental cyber-epidemic model to analyze the spread of malicious code within computer networks. The model classifies network nodes into different states: susceptible (S), exposed (E), infected (I), controlled (C), recovered (R), protected (P), disabled (D), and attackers (A). It captures important aspects of a cyberattack, such as the spread of infection, botnet formation, security measures, and achieving a state of destruction within a compromised system. Basic reproduction number \((R_0)\), is derived in order to measure the systematic stability and determine whether cyber threats will persist or be mitigated. Equilibrium analysis is conducted to determine parameters under which condition malware free equilibrium is globally asymptotically stable, and the endemic equilibrium remain locally stable. Moreover, (\( R_0 \)) is used to carry out a sensitivity analysis to check the impact of the parametric values in the system. Our proposed model is solved via numerical simulation using the Runge-Kutta-Fehlberg (RKF45) method to gain an understanding of network security, malware spread, security measures needed, and the most cost efficient solutions. Results verified through MATLAB simulations, align with real-world cyberattack patterns, offering practical implications for securing modern network infrastructures.