加密环境下大数据特征集并行存储方法研究  

Research on parallel storage method of big data feature set in encrypted environment

作　　者：李蓉蓉[1] LI Rongrong(Guangdong University of Science and Technology,Dongguan 523000,China)

机构地区：[1]广东科技学院

出　　处：《现代电子技术》2019年第18期69-72,共4页Modern Electronics Technique

基　　金：广东省自然科学基金项目(E17094);广东省高校重大科研项目(青年创新人才类项目)(2017KQNCX226)~~

摘　　要：针对传统方法存在数据存储速度慢,容量小的问题,提出一种加密环境下大数据特征集并行存储方法.采用MMSE算法对大数据特征集进行优化处理,以获得并行的大数据特征集.在加密环境下,当内存中保留数据记录达到一定数量时,在并行处理后的数据中插入批量Hash索引.以此为基础,基于Map函数对数据进行映射处理,进而完成对大数据特征集的并行存储.实验结果表明,利用所提方法的存储过程中数据分布通道具有较好的一致性,且存储速度约为传统方法的3倍,存储容量大.A parallel storage method of big data feature set in encryption environment is proposed to solve the problems ex- isting in traditional methods for its slow data storage speed and small capacity. The MMSE algorithm is used to optimize the big data feature set to obtain the parallel big data feature set. In the encrypted environment,the batch Hash index is inserted into the parallelly-processed data when the data records retained in memory reaches a certain quantity. According to it,the data map- ping processing is conducted based on Map function,and then the parallel storage of big data feature set is completed. The ex- perimental results show that the data distribution channel has better consistency in the storage process by means of the proposed method,the storage speed of the proposed method is as about 3 times as that of the traditional method,and the storage capacity is large.

关 键 词：大数据特征集 并行存储 MMSE算法 Hash索引 结果分析 映射处理 

分 类 号：TN918.4.34[电子电信—通信与信息系统] TP311.13[电子电信—信息与通信工程]