Prediction model of network security situation based on regression analysis

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.