A Chaotic Characteristics Identification Method for Network Security Situation Time Series

摘要

To choose a reasonable prediction model of network security situation, an Identification Method of Chaotic Time Series (IMCTS) used for identifying chaotic characteristics of network security situation time series is proposed. In IMCTS, the C-C method is used for calculating embedding dimension τ and delay time d of time series; the Fast Fourier Transform (FFT) method is used for calculating mean period P of time series; the small data sets method is used for obtaining largest Lyapunov exponent λ_1 of time series, and then λ_1 is used for identifying chaotic characteristics of time series so as to choose a reasonable time series prediction model. To identify chaotic characteristics of time series, network security situation time series obtained from DARPA 1999 Data Set are used for simulation experiments. The experimental results show that IMCTS provides a strong theoretical basis for choosing an effective prediction model of network security situation, and meanwhile holds broad application prospect.