METHOD FOR DETERMINING THE EFFECTIVENESS OF LOCAL NETWORK SECURITY SYSTEM

Abstract

A mathematical model has been developed and the model of local network protection has been studied. The dependencies of the information flow in the local network on the components of information protection, system security, the size of the threat system to information security, and the network clustering coefficient have been considered.

A system of linear equations has been obtained, which reflects the impact of the information flow change rate on the security of the local network and coefficients that characterize: the impact of security measures, such as protection against electromagnetic radiation, physical security, data integrity and authenticity control, access control to information, firewall operation, antivirus protection, software and hardware component failures, the impact of personal data leakage rate, the amount of personal data, user identification and authentication, data backup, auditing, the effect of security on information leakage, and the effect of system size on security.

As a result of solving the system of differential equations, mathematical and graphical dependencies of the personal data protection indicator in the local network on various components have been obtained.

By considering three variants of solving the equation near the steady-state of the system, it was concluded that, depending on the ratio of dissipation and the system's own oscillation frequency, the decay of the amplitude to a certain value may occur periodically with a decaying amplitude or according to an exponentially decaying law. A more visual analysis of the system's behavior has been performed by transitioning from the differential form of the equations to the discrete form and modeling a certain interval of the system's existence.

Mathematical and graphical dependencies of the system's natural frequency, oscillation period, and damping coefficient are presented.

Simulation modeling has been carried out for values deviating from the system's steady-state. The simulation results demonstrate that the local network protection system is nonlinear.