Automatic cyber security risk assessment based on fuzzy fractional ordinary differential equations

Abstract Real-time network security risk detection is of great significance to the research of dynamic network security, and it is one of the current research hotspots of network security. Based on the idea of artificial immune, a dynamic network intrusion detection and prediction model based on fuzzy fractional ordinary differential equations is proposed. The uniqueness of the solution is studied in the square integrable equation space using the principle of compression operator, and then the second type of solution is proposed. The differential-type piecewise Taylor series expansion method of the linear Fredholm integral equation is used to obtain the approximate solution expressions and perform convergence and error estimation. The process of network attack detection, antibody concentration calculation method and risk prediction process based on time series autoregressive moving average model are given. The experimental results show that the model can quantitatively analyze the current security situation of the network in real time and make predictions about the risks faced by the network. The prediction effect of catastrophe network risk is better than the GM (1, 1) model, and it is close to the actual risk situation, with high prediction accuracy.