Network information security protection method based on additive Gaussian noise and mutual information neural network in cloud computing background

Abstract

In the cloud computing environment, data security and privacy have received unprecedented attention, but current information security protection methods cannot simultaneously balance data utility and privacy protection effects. Therefore, a network information security protection method based on Gaussian denoising and mutual information neural network is proposed. The research aims to protect network information while maintaining high data utility. This study utilizes Gaussian noise and K-dimensional perturbation trees to establish a privacy protection scheme, and introduces a Bayesian network-based network intrusion detection method to combine the two for information privacy protection. Afterwards, mutual information is used to evaluate the effectiveness of privacy protection and further optimize the parameters of the protection scheme. The experimental results showed that the proposed method achieved a data utility retention rate of 85%, and the number of privacy breaches did not exceed 3 times. In long-term experiments, through continuous optimization, the number of breaches gradually remained at 0. From this, the proposed privacy protection method can effectively improve the data security and privacy in cloud computing environments, and ensure data utility during transmission and storage processes.