云计算中数据存储的完整性校验模型研究

发布时间：2018-02-27 13:23

本文关键词： 云存储数据完整性验证 Counting Bloom Filter　出处：《东华大学》2013年硕士论文　论文类型：学位论文

【摘要】：随着云计算的迅速发展,越来越多的数据被存放在云服务器中。云存储作为一种新型的存储服务,所具有的高可扩展性、高可靠性、价格低廉和可随时随地接入访问的特点,为海量数据的存储管理提供了全新的模式。但是,由于用户把数据都托管给了云服务商,失去了对数据的本地监管控制权,因此面临着很多挑战。其中一个重要的问题,就是用户如何确定存放在云服务器中的数据是完整的。传统的数据完整性校验方法需要将数据下载到本地才能进行验证。由于云存储中数据量极大,下载数据块将带来很大的网络负担,因此传统的方法和技术并不适用于云存储环境。目前,已有的针对云存储中数据完整性校验算法中以基于RSA公钥技术的算法较为优秀,它们以数据块为基本校验单元,具有无限次校验、校验时隐私保护、通信开销低等特性。但由于算法中大量模指数运算,其计算开销很大。且考虑到各方开销,算法中数据分块大小受到较为严格的限制。考虑到云存储中数据的动态性,即数据块在存储期间会多次发生修改、删除等更新操作,使得原来计算的校验元失效而频繁更新,进一步加重计算开销。本文在支持第三方验证的基础上,提出了一种基于Counting Bloom Filter(CBF)的数据完整性校验算法,用于高效地校验云存储中动态数据的完整性。由于数据的动态性使得同一块数据校验无需进行无限次,算法通过使用速度很快的哈希技术,避免使用基于RSA公钥技术算法中大量模指数运算,有效降低了计算开销。同时,本文利用CBF空间效率高的特点,提出了概率性的校验元,以很小的概率损失实现了校验值的压缩,节约了存储和通信开销。另外,相对于基于RSA公钥技术的算法,本算法中数据分块大小对整体开销的影响大大降低。最后通过理论分析和模拟实验表明,算法在数据动态变化环境下,以较低的存储空间代价实现了在文件生命周期内轻量级的完整性校验,尤其是在挑战-应答过程中,降低了云服务器和第三方验证方在计算方面的开销。
[Abstract]:With the rapid development of cloud computing, more and more data are stored in cloud servers. As a new storage service, cloud storage has the characteristics of high scalability, high reliability, low price and easy access. Provides a new model for the storage and management of massive data. However, because users host data to cloud service providers and lose local regulatory control over the data, they face many challenges. Is how the user determines that the data stored in the cloud server is complete. Traditional data integrity verification methods need to download the data to the local to verify. Downloading data blocks will bring great network burden, so traditional methods and techniques are not suitable for cloud storage environment. At present, the existing algorithms based on RSA public key technology for data integrity verification in cloud storage are excellent. They take the data block as the basic check unit, have the characteristics of infinite check, privacy protection during checking, low communication overhead, etc. However, due to a large number of modular exponent operations in the algorithm, the computation cost is very large. The size of the data block in the algorithm is strictly limited. Considering the dynamic nature of the data in cloud storage, that is, the updating operations such as data blocks will be modified and deleted many times during storage, which makes the original calculated checkpoints fail and update frequently. In this paper, a data integrity verification algorithm based on Counting Bloom filter is proposed based on the support of third party verification. It is used to verify the integrity of dynamic data in cloud storage efficiently. Because of the dynamic nature of data, the same block of data need not be checked infinitely. In order to avoid the use of a large number of modular exponent operations in the algorithm based on RSA public key technology, the computational overhead is effectively reduced. At the same time, a probabilistic check element is proposed in this paper, which makes use of the high efficiency of CBF space. In addition, compared with the algorithm based on RSA public key technology, the compression of check value is realized with very small probability loss, and the storage and communication overhead is saved. In this algorithm, the effect of data block size on the overall overhead is greatly reduced. Finally, theoretical analysis and simulation experiments show that the algorithm is in the dynamic environment of data change. Lightweight integrity verification is implemented in the file lifecycle at a lower storage cost, especially in the challenging response process, which reduces the computational overhead of cloud servers and third-party verifiers.
【学位授予单位】：东华大学
【学位级别】：硕士
【学位授予年份】：2013
【分类号】：TP333

【参考文献】