可撤销属性加密结合快速密度聚类算法的非结构化大数据安全存储方法  

A SECURE STORAGE METHOD FOR UNSTRUCTURED BIG DATA USING REVOCABLE ATTRIBUTE ENCRYPTION COMBINED WITH FAST DENSITY CLUSTERING ALGORITHM

作　　者：谷保平 马建红[2] Gu Baoping;Ma Jianhong(School of Information Engineering,Zhengzhou Vocational University of Information and Technology,Zhengzhou 450046,Henan,China;School of Software,Zhengzhou University,Zhengzhou 450002,Henan,China)

机构地区：[1]郑州信息科技职业学院信息工程学院,河南郑州450046 [2]郑州大学软件学院,河南郑州450002

出　　处：《计算机应用与软件》2021年第5期337-343,共7页Computer Applications and Software

基　　金：河南省科技厅科技攻关计划项目(172102210565,182102210576)。

摘　　要：针对非结构化大数据难以实现安全存储和易遭受安全攻击的问题,提出可撤销属性加密结合快速密度聚类算法的非结构化大数据安全存储方法。利用可撤销属性方法为非结构化大数据提供安全的存储结构,通过区分安全攻击和传输错误来防止大数据的误传和避免安全攻击;利用霍夫曼压缩技术对数据进行快速压缩,节省非结构化大数据处理过程中的时间开销;利用错误控制技术为潜在丢失的数据提供备份系统,并利用快速密度聚类算法有效处理多维大数据文件。实验证明,相比于其他现有非结构化大数据安全存储方法,该方法的执行速度更快,时间开销更小,信息损失百分比更低,信噪比(SNR)和压缩比更高。Aiming at the problems of slow classification speed and vulnerability to security attacks,a secure storage method of unstructured big data based on revocable attribute encryption combined with fast density clustering algorithm is proposed.The retractable attribute method was used to provide a secure storage structure for unstructured big data.The security of data transmission and storage was improved by distinguishing security attacks and transmission errors to prevent the mis-transmission of big data and avoid security attack.Then the Hoffman compression technology was used to quickly compress data,which saved time overhead in the processing of unstructured big data.Finally,an error control technology was used to provide a backup system for potential lost data,and a fast density clustering algorithm was used to process multi-dimensional big data files effectively.The experiments show that compared with other existing security storage methods of unstructured big data,the proposed security storage method of unstructured big data has been verified to have faster execution speed,lower time cost,lower percentage of information loss,higher SNR and compression ratio.

关 键 词：可撤销属性加密 快速密度聚类算法 非结构化 大数据 安全存储 安全攻击 霍夫曼压缩 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]