融合信誉评价的云存储安全优化方法研究

发布时间：2018-01-07 01:09

本文关键词：融合信誉评价的云存储安全优化方法研究　出处：《河南科技大学》2014年硕士论文　论文类型：学位论文

【摘要】：云计算（Cloud computing）的基础是虚拟化，通过互联网提供整合的基础设施、平台、软件等服务，将分布式的计算资源、存储等聚合成具有超强的计算能力的虚拟计算机。云存储（Cloud storage）作为云计算的延伸概念，为网络应用在线存储的海量数据提供了有效的解决方案，在节约存储资源投入的情况下，提供了一个远端备份、在线存储访问的大型数据中心。但云存储服务目前面临许多安全问题，用户向云端存放敏感数据时，由于无法直接控制存储资源，，对云端数据可能发生的安全威胁无法感知，对云存储服务难以信赖。本文针对云存储服务可靠性的问题，在分析了现有的云存储安全技术的基础上，融合信誉机制为云存储服务提供可靠地信誉保障，增加可信第三方为云存储服务参与方提供安全监测，并通过优化冗余灾备保障云端数据高可用性。主要研究工作包括以下几个方面：（1）分析了现有的云存储关键技术，全面系统的阐述了现有研究进展，介绍了目前云存储中存在的主要问题，讨论当前相关技术的研究方法。（2）构建云存储服务信誉管理机制，一方面针对云服务数据量大、强动态性、集中式管理复杂的问题，融合自律计算的思想，提出了一种自主信誉管理机制。另外，针对不实评价进行信誉修正，提出了一种不实评价过滤方法。（3）提出了一种融合信誉评价的云存储模型，使用户可以从提供商中获得高质量的可靠存储服务，同时在用户和服务提供商间建立一种直接的互相信任。提出的云存储模型支持数据块级的动态操作、支持授权用户访问外包数据的最新版本、允许数据属主授予或者撤销外包数据的访问权限。（4）针对分布在众多弹性节点中的数据高可用性保障，降低数据冗余度，提出了一种基于擦除码的高可靠云存储方案，通过编码冗余保证数据文件高可用性同时优化存储冗余度，利用Chord环提升数据块分布式容灾存储检索效率。
[Abstract]:Cloud computing is based on virtualization, providing integrated infrastructure, platforms, software, and other services over the Internet, and distributed computing resources. Storage is aggregated into a virtual computer with super computing power. Cloud storage is an extended concept of cloud computing. It provides an effective solution for the massive data stored online by network applications, and provides a remote backup under the condition of saving the investment of storage resources. However, cloud storage service is faced with many security problems. When users store sensitive data to the cloud, they can not directly control storage resources. You can't perceive security threats to cloud data and trust cloud storage services. Aiming at the problem of cloud storage service reliability, based on the analysis of the existing cloud storage security technology, this paper integrates the reputation mechanism to provide a reliable reputation guarantee for cloud storage service. Adding trusted third parties to provide security monitoring for cloud storage service participants and ensuring high availability of cloud data by optimizing redundant disaster preparedness. In this paper, the key technologies of cloud storage are analyzed, the current research progress is described systematically, the main problems existing in cloud storage are introduced, and the research methods of related technologies are discussed. Build cloud storage service reputation management mechanism, on the one hand for cloud service data volume, strong dynamic, centralized management of complex problems, the integration of the idea of self-discipline computing. An independent reputation management mechanism is proposed, and a filtering method for false evaluation is proposed. A cloud storage model integrating reputation evaluation is proposed to enable users to obtain high quality reliable storage services from providers. At the same time, a direct mutual trust between users and service providers is established. The proposed cloud storage model supports block level dynamic operation of data, and supports authorized users to access the latest version of outsourced data. Allow the data to be granted or revoked by the master. In order to ensure the high availability of data and reduce the data redundancy, a high reliability cloud storage scheme based on erasure code is proposed. The high availability of data files and the optimization of storage redundancy are ensured by encoding redundancy, and the efficiency of distributed disaster recovery storage retrieval of data blocks is improved by using Chord loop.
【学位授予单位】：河南科技大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TP333

【参考文献】