基于全同态加密优化的云数据隐私保护方法
Cloud Data Privacy Protection Method Based on Homomorphic Encryption Optimization
王雪飞 1王鹏 2佟良3
作者信息
- 1. 绥化学院电气工程学院,黑龙江 绥化 152061
- 2. 绥化学院信息工程学院,黑龙江 绥化 152061
- 3. 黑龙江大学数学科学学院,黑龙江 哈尔滨 150080
- 折叠
摘要
为了提升云数据的安全性,改善数据加密时间长以及加密效果较差等问题,提出一种基于全同态加密优化的云数据隐私保护方法.通过生成对抗网络模型学习原始数据中的重要特征,确保合成的数据和初始数据之间具有较高的相似度,使其能够满足差分隐私特征.引入PKI对数据所有者和用户身份认证处理,完成密钥的产生和分发,根据差分隐私特征将用户所有者的数据划分为不同类型,获取用户访问特征向量.构建全同态加密机制,同时引入代理重加密机制,通过密钥转换完成数据加密处理,最终实现云数据全同态加密优化.实验对比结果表明,上述加密方案能够快速完成数据加密处理,有效确保数据的安全性.
Abstract
In order to improve the security of cloud data and the data encryption,a method of protecting cloud da-ta privacy based on homomorphic encryption optimization was put forward.Firstly,the model of generative adversarial network was used to learn important features in the original data,thus ensuring a high similarity between synthesized data and initial data.In this way,it can meet the differential privacy characteristics.Then,PKI was introduced to au-thenticate the data owner and user identity,thus completing the generation and distribution of the key.According to the differential privacy characteristic,the data of owner was divided into different types to obtain the user access fea-ture vector.Moreover,an all-homomorphic encryption mechanism was established.Meanwhile,an agent re-encryption mechanism was introduced.After that,the data encryption through key conversion was completed.Finally,we realized the optimization of all-homomorphic encryption of cloud data.The experimental results show that the proposed en-cryption scheme can quickly complete data encryption,thus effectively ensuring the security of data.
关键词
云数据/隐私保护/全同态加密算法/对抗网络生成/代理重加密机制Key words
Cloud data/Privacy protection/All homomorphic encryption algorithm/Generative adversarial net-work/Agent re-encryption mechanism引用本文复制引用
基金项目
绥化学院科研启动基金(SQ21007)
黑龙江省科研业务费科研项目(YWK10236200136)
黑龙江省科研业务费科研项目(KYYWF10236180107)
出版年
2024
NETL
NSTL
万方数据