可证明数据持有模型与方案的研究

发布时间：2018-11-15 10:46

【摘要】：随着数据爆炸和宽带网络的普及，云存储已经成为当今云计算领域的一个重要应用分支，越来越多的用户趋向于将自己的数据存储到云端。然而，存储在云端的数据依然存在设备故障、黑客攻击、恶意篡改等威胁，云存储服务方需要通过一种有效的协议来向用户提供数据完整的证明。可证明数据持有(PDP)提供了一个解决上述问题的模型，在该模型下，用户和服务方之间通过一种挑战-应答协议来验证数据的完整性。在此基础上，公开验证和动态更新这两大特性扩充了PDP模型的适用范围。然而，这些特性的引入不仅增加了方案设计的复杂度，也对效率产生了很大影响：如对数据块标签和动态更新均不能实现批量验证，同时公开验证下用户数据存在被第三方验证机构窃取的可能，，用户隐私无法得到保障。文本针对上述问题，对支持公开验证和动态更新的可证明数据持有模型和方案做了如下改进：基于树形hash认证结构，提出了在公开验证下数据块标签的批量验证方案，从而提升静态标签的验证效率；改进现有基于单个数据块更新的动态PDP模型，提出了批量数据块更新的思想。设计了用于支持批量更新认证的BA23-Tree及相应的批量更新和验证算法。仿真实验表明，在目标数据块均匀分布情况下，批量更新效率可以稳定提升一倍左右；而在目标数据块连续分布的情况下，更新效率将随批大小的增大而显著提升；在公开验证下，完善现有PDP的安全模型，给出对消息进行盲化处理的方案，使得在不泄露用户原始数据内容的前提下，保证第三方验证机构依旧有能力对数据进行完整性校验。
[Abstract]:With the popularization of data explosion and broadband network, cloud storage has become an important application branch in cloud computing field. More and more users tend to store their data to the cloud. However, the data stored in the cloud still has some threats, such as device failure, hacker attack, malicious tampering and so on. Cloud storage service needs to provide the user with complete proof of data through an effective protocol. The provable data holding (PDP) provides a model to solve the above problem, in which the integrity of the data is verified by a challenge-reply protocol between the user and the service party. On this basis, open validation and dynamic update extend the applicability of PDP model. However, the introduction of these features not only increases the complexity of the scheme design, but also has a great impact on the efficiency: for example, the data block label and dynamic update can not achieve batch verification. At the same time, there is the possibility that user data can be stolen by a third party certification authority under public verification, and user privacy cannot be protected. Aiming at the above problems, the paper improves the provable data holding model and scheme which supports public verification and dynamic updating. Based on the tree hash authentication structure, a batch verification scheme of data block label under public verification is proposed. In order to improve the verification efficiency of static tags; This paper improves the existing dynamic PDP model based on single data block updating and proposes the idea of batch data block updating. The BA23-Tree and the corresponding batch update and verification algorithm are designed to support batch update authentication. The simulation results show that under the condition of uniform distribution of the target data block, the batch update efficiency can be increased by about double steadily, but when the target data block is distributed continuously, the updating efficiency will be improved significantly with the increase of the batch size. Under public verification, the security model of existing PDP is improved, and the scheme of blind message processing is given, so that the original data content of the user is not disclosed. Ensure that the third-party certification authority still has the ability to verify the integrity of the data.
【学位授予单位】：上海交通大学
【学位级别】：硕士
【学位授予年份】：2013
【分类号】：TP333;TP309

【相似文献】