Network Security Situation Prediction Using Artificial Immune System and Phase Space Reconstruction

Abstract

To solve network security situation prediction problem, a novel Prediction approach for network security situation based on artificial Immune system and Phase Space Reconstruction (PIPSR) is proposed. In PIPSR, we use phase space reconstruction to analyze the time series of network security situation and to reconstruct proper time series phase space; then, we use immune evolution mechanism to construct corresponding prediction model for network security situation; and lastly we use this prediction model to forecast network security situation. The simulation results show that PIPSR can more exactly forecast the future network security situation than GAP and BPNNP.