通信学报2024，Vol.45Issue(3)：50-65,16.DOI:10.11959/j.issn.1000-436x.2024050
基于函数加密的密文卷积神经网络模型
Convolutional neural network model over encrypted data based on functional encryption
摘要
Abstract
Currently,homomorphic encryption,secure multi-party computation,and other encryption schemes are used to protect the privacy of sensitive data in outsourced convolutional neural network(CNN)models.However,the computa-tional and communication overhead caused by the above schemes would reduce system efficiency.Based on the low cost of functional encryption,a new convolutional neural network model over encrypted data was constructed using functional encryption.Firstly,two algorithms based on functional encryption were designed,including inner product functional en-cryption and basic operation functional encryption algorithms to implement basic operations such as inner product,mul-tiplication,and subtraction over encrypted data,reducing computational and communication costs.Secondly,a secure convolutional computation protocol and a secure loss optimization protocol were designed for each of these basic opera-tions,which achieved ciphertext forward propagation in the convolutional layer and ciphertext backward propagation in the output layer.Finally,a secure training and classification method for the model was provided by the above secure pro-tocols in a module-composable way,which could simultaneously protect the confidentiality of user data as well as data labels.Theoretical analysis and experimental results indicate that the proposed model can achieve CNN training and clas-sification over encrypted data while ensuring accuracy and security.关键词
卷积神经网络/密文数据/函数加密/隐私保护Key words
convolutional neural network/encrypted data/functional encryption/privacy protection分类
信息技术与安全科学引用本文复制引用
王琛,李佳润,徐剑..基于函数加密的密文卷积神经网络模型[J].通信学报,2024,45(3):50-65,16.基金项目
国家自然科学基金资助项目(No.62372096,No.62173101) The National Natural Science Foundation of China(No.62372096,No.62173101) （No.62372096,No.62173101）
