Prediction model of network security situation based on regression analysis

Abstract

This paper analyses and compares the existing network security situation evaluation methods, defines various network security evaluation concepts, and proposes a method for network security situation evaluation using virtual Honeynet. With the relationships among Honeynet active, host active of computer networks and IP active when network intrusion occurs, a binary linear regression prediction model is proposed. In the end, a prototype system is designed for data collection and regression fitting, and the validity of the regression prediction model is proved.