基于密度划分的云数据分块存储方法仿真  

Simulation of Cloud Data Block Storage Method Based on Density Division

作　　者：潘文标 元文浩 PAN Wen-biao;YUAN Wen-hao(Information Technology Center Wenzhou Medical University,Wenzhou Zhejiang 325035,China)

机构地区：[1]温州医科大学信息技术中心,浙江温州325035

出　　处：《计算机仿真》2022年第8期456-459,495,共5页Computer Simulation

基　　金：市级课题:云计算视域下的虚拟资源分配优化策略(R20210095)。

摘　　要：为减小云数据存储冗余度、缓解存储压力,面向细粒度云数据类型,设计出一种基于密度划分算法的分块存储方法。根据类别相似度与跨类别数据属性,利用密度划分算法聚类云数据。分析聚合条件的判定矩阵与距离阈值,取得两参数最优值。将各类别云数据划分为规格相同的数据块,采用里所码二次分块,以伽罗华域为运算平台,通过范德蒙矩阵编码、解码处理,在多个适配度较高的节点上,完成细粒度云数据块存储。仿真结果表明,所提方法的聚类精度较高,可用性较好;证明所提方法具有显著的优越性与良好的实践性。In order to reduce the redundancy of cloud data storage and relieve the stress of storage,this article presented a block storage method for fine-grain cloud data based on density division.According to category similarity and cross-category data attributes,we used the density division algorithm to cluster cloud data,and then analyzed the decision matrix and distance threshold of the clustering condition,and thus to obtain the optimal values of the two parameters.After that,we divided all cloud data into blocks with the same specifications and used Reed-Solomon code to divide them again.Taking Galois Field as the computing platform,we used the Vandermonde matrix for encoding and decoding the data block.Finally,we completed the fine-grain cloud data block storage on multiple nodes with higher adaptability.Simulation results prove that the proposed method has high clustering accuracy and good usability.Consequently,this method has significant advantages and good practicality.

关 键 词：密度划分 细粒度云数据 分块存储 范德蒙矩阵编码 

分 类 号：TP399[自动化与计算机技术—计算机应用技术]