Network security situation awareness model and method based on CE-RBF

Abstract

The invention discloses a network security situation awareness model and method based on CE-RBF. The model comprises a data preprocessing module, a situation calculation module, a parameter optimization module and a situation prediction module. The method comprises the following steps: collecting data sets from different sources, and extracting principal component information for situation awareness to obtain asset attack threat data and system state data; calculating a risk value according to the asset attack threat data of the network equipment, and evaluating the security situation of the whole network; determining initial parameters of the RBF neural network, establishing an optimization objective function, optimizing the parameters in the optimization objective function by using a CEalgorithm, substituting the optimal parameter set into the RBF neural network after finding the optimal parameter set, and training by using historical network situation values as sample data; and performing situation prediction by using the trained RBF neural network. According to the method, the problem of parameter optimization in the high-dimensional model is solved by utilizing the efficientoptimization capability of CE, and the prediction capability of the neural network is improved.

Description

technical field [0001] The invention relates to the field of network information security, in particular to a CE-RBF-based network security situation awareness model and method. Background technique [0002] With the rapid development of network technology, network attack incidents are also increasing year by year. Network security issues have become the focus of people's current attention. Forecasting the network security situation can grasp the security status of the network before the network attack incident occurs, so as to take corresponding protection. Measures to avoid unnecessary attacks and losses. [0003] The research on network security situational awareness started relatively late in China, and most of them focus on the research of network threat quantification process and intrusion detection process, while a few prediction models can only be applied to specific standard systems and application scenarios. As an efficient feed-forward neural network, RBF (Radial...

AI Technical Summary

Problems solved by technology

[0004] For the defects existing in the prior art, the object of the present invention is: by combining CE (Covariance Matrix Adaptation Evolutionary Strategies, covariance matrix adaptive evolution strategy, or claiming CMA-ES) algorithm and RBF neural network model, realize a An efficient and accurate network security situation awareness model and its perception method. The model calculates the network security situation value by classifying and counting a large number of vulnerability information and log information, and uses it as the input of the RBF neural network, and uses the CE algorithm to optimize the neural network. The parameters of the network can improve its generalization ability, accurately predict the network security situation value in the future time period, and draw the network security situation prediction map, so as to solve the problem of inaccurate prediction effect of RBF neural network

Images