A prediction model of cloud security situation based on evolutionary functional network
- PDF / 2,788,936 Bytes
- 15 Pages / 595.276 x 790.866 pts Page_size
- 6 Downloads / 181 Views
A prediction model of cloud security situation based on evolutionary functional network Baowen Xie 1 & Guosheng Zhao 1 & Mianxing Chao 1 & Jian Wang 2 Received: 6 July 2019 / Accepted: 27 December 2019 # Springer Science+Business Media, LLC, part of Springer Nature 2020
Abstract Aiming at the dynamic uncertainty and prediction accuracy of security situation prediction in complex cloud network environment, a prediction model of cloud security situation based on evolutionary functional network is proposed. Firstly, the evolutionary functional network model is constructed by combining the evolutionary algorithm with the functional network, which solves the problem of basis function selection and basis function coefficient correction of the prediction model. Secondly, the stochastic approximation algorithm is used to process and comprehend the cloud security situation elements, and the computational complexity of the prediction model is reduced by the dimensionality reduction method. Finally, by constructing the credibility matrix of the uncertain influence relationship of security situation elements, we use the multivariate non-linear regression algorithm to predict the cloud security situation. The simulation results show that compared with BP model and RAN-RBF model, the prediction accuracy of the proposed model is improved by 8.2% and 6.9% respectively, and the convergence efficiency is improved by 12.3% and 10.8% respectively. Keywords Cloud security . Situation prediction . Evolutionary functional network . Multivariate nonlinear regression
1 Introduction In recent years, cloud computing as an IT infrastructure has been widely and deeply applied in various fields. Its unprecedented openness and complexity propose the new challenges to guarantee its security. Cloud security is the primary factor that affects the rapid development of cloud computing. Cloud security situational awareness is a process of intelligent reasoning for the optimal decision-making of cloud security and the optimization of cloud security management. Cloud security situational awareness includes cloud security situation elements extraction, situation comprehension and situation prediction [1]. The prediction of cloud security situation is an important part of cloud security situation awareness. The prediction of cloud security situation can realize the dynamic
* Guosheng Zhao [email protected] 1
College of Computer Science and Information Engineering, Harbin Normal University, Harbin 150025, China
2
School of Computer Science and Technology, Harbin University of Science and Technology, Harbin 150001, China
management of cloud network security and prevent largescale cloud security events. At present, there are many research methods, such as rough set prediction, time series prediction and neural network prediction. In order to improve the quality and efficiency of extracting security situation elements, Zhao et al. [2] proposed a parallel reduction algorithm based on attribute importance matrix. The idea of parallel reduction is introduced
Data Loading...
