云存储中数据完整性验证机制的研究与实现

发布时间：2018-03-02 03:37

本文关键词： 云存储安全数据完整性验证认证字典 hadoop　出处：《电子科技大学》2013年硕士论文　论文类型：学位论文

【摘要】：云存储技术使得存储服务变得更加便利，其作为一种兼具实用价值与市场价值的服务运营模式，已经引起相关人士越来越多的关注。云存储服务的思路与理念对于无论是企业用户还是个人用户都具有吸引力，但是正如其他的存储服务模式一样，其数据安全的问题将是用户的关注和担忧所在。另外云存储自身的技术特点使得其数据安全问题比传统模式更为复杂，因此在云存储服务由概念走向市场的进程中，保障云存储中数据的安全是至关重要的一个课题。本文主要针对云存储中数据的完整性验证机制进行研究。云存储中的数据完整性验证主要分为两个方面：远程完整性验证与一般完整性验证。远程完整性验证是指使用挑战/响应模式来对数据完整性进行概率性质的验证，而一般完整性验证是基于访问的，它应用认证字典技术来对访问的数据进行完整性的验证。本文主要着眼于对一般完整性验证机制开展研究，分析了云存储的三方存储模型，并在现有的数据完整性验证机制方面的相关研究基础上，提出并设计两种新的更加优化的数据完整性验证方案。本文的主要内容有：（1）提出了基于B+树的数据完整性验证方案，该方案不但在验证过程中具有较低的时间复杂度，同时更加灵活地支持了数据的动态更新操作，，其中单向累加器技术的应用使得数据更新时各结点摘要值的变化更加高效。（2）提出了基于全结点存储树的数据完整性验证方案，并以结点大小平衡树为例进行方案的设计与分析。这种全结点存储树节约了认证字典的存储空间，同时兼具平衡二叉搜索树的平衡性和高效性，是一种全新的认证字典构造方式。（3）基于Hadoop技术中的分布式文件系统（HDFS）建立一个云存储的原型系统，实现文件数据的上传、下载、删除等操作，并为本文所提出的完整性验证方案提供平台支持。另外，还利用密码学方法实现了B+树以及SBT结构的可认证化。
[Abstract]:Cloud storage technology makes storage service more convenient, as a service operation mode with both practical value and market value. Cloud storage services are attractive to both enterprise and individual users, but just like other storage service models, In addition, the technical characteristics of cloud storage make its data security problem more complex than the traditional model, so in the process of cloud storage service from concept to market, It is very important to ensure the security of data in cloud storage. Data integrity verification in cloud storage is mainly divided into two aspects: remote integrity verification and general integrity verification. Remote integrity verification is mainly divided into two aspects: remote integrity verification and general integrity verification. Is the use of challenge / response patterns to verify the probabilistic nature of data integrity, General integrity verification is based on access, which uses authentication dictionary technology to verify the integrity of accessed data. This paper focuses on the general integrity verification mechanism, analyzes the three-party storage model of cloud storage, and on the basis of the existing data integrity verification mechanism, Two new and more optimized data integrity verification schemes are proposed and designed. The main contents of this paper are as follows:. 1) A data integrity verification scheme based on B-tree is proposed. This scheme not only has lower time complexity in the verification process, but also supports the dynamic update operation of data more flexibly. The application of unidirectional accumulator makes the change of summary value of each node more efficient. (2) A data integrity verification scheme based on the all-node storage tree is proposed, and the scheme is designed and analyzed by taking the node size balance tree as an example. This kind of all-node storage tree saves the storage space of the authentication dictionary. At the same time, it has the balance and efficiency of balanced binary search tree, so it is a new way to construct authenticated dictionary. Based on the distributed file system (HDFS) in Hadoop technology, a prototype system of cloud storage is established, which can upload, download and delete file data, and provide platform support for the integrity verification scheme proposed in this paper. The B tree and SBT structure are also authenticated by cryptography.
【学位授予单位】：电子科技大学
【学位级别】：硕士
【学位授予年份】：2013
【分类号】：TP333

【相似文献】