混合云环境下多维数据安全去重算法研究  

Research on Security Deduplication Algorithm of Multidimensional Data in Hybrid Cloud Environment

作　　者：杨丽华 鄂晶晶 冯锋 YANG Li-hua;E Jing-jing;FENG Feng(Computer School,Hulunbuir College,Inner Mongolia Hulunbuir 021000,China;School of Information Engineering Ningxia University,Yinchuan Ningxia 750021,China)

机构地区：[1]呼伦贝尔学院计算机学院,内蒙古呼伦贝尔021000 [2]宁夏大学信息工程学院,宁夏银川750021

出　　处：《计算机仿真》2023年第12期543-547,共5页Computer Simulation

基　　金：宁夏自然科学基金重点资助项目(2021AAC02004)。

摘　　要：多维数据重复率过高,即多维数据冗余杂乱,会导致目标数据检索困难,数据计算复杂度提升,为了降低多维数据冗余度,提高目标数据检索效率,提出混合云环境下多维数据安全去重算法。利用小波变换和高白光干涉信号对原始多维数据去噪,根据局部非参数建模方法提取去噪后的多维数据特征;运用Simhash算法和聚类算法,获取多维数据特征的可识别镜像分段聚类结果;利用DBSCAN算法分割以相同镜像分段为单位的同组多维数据,实现整体多维数据的安全去重。仿真结果表明,所提方法计算时间开销少、去重率高。The high repetition rate of multidimensional data,i.e.the redundancy and disorder of multidimensional data,can lead to difficulties in target data retrieval and increased data computational complexity.In order to reduce the redundancy of multidimensional data,this paper improves the efficiency of target data retrieval and proposes a multidimensional data security deduplication algorithm in a hybrid cloud environment.Foremost,wavelet transform and high white light interference signal were used to remove noise from original multidimensional data.And then,fea⁃tures of denoised multidimensional data were extracted by using the local nonparametric modeling method.Moreover,Simhash algorithm and clustering algorithm were adopted to obtain the piecewise clustering results of recognizable im⁃ages of multi-dimensional data features.Finally,DBSCAN algorithm was used to segment the multidimensional data in the same group and same image segment.Thus,the safe deduplication for multidimensional data was achieved.Simulation results show that the proposed method needs less computing time,with high deduplication rate.

关 键 词：多维数据 小波变换 特征提取 聚类算法 

分 类 号：TP309.3[自动化与计算机技术—计算机系统结构]