Analysis And Design Of Security Auditing Scheme In Cloud Computing

As one of the important trends of the current information technology, Cloud computing makes individuals or all types of enterprises users with convenient for storage, relatively cheap price and computing resources through remote data access. Cloud storage, as an important branch of cloud computing, makes cloud user’s data be stored on a remote cloud server. In cloud storage, the users lose the physical storage of their data, which results the security of the data only ensured unilaterally by the server. In recent years, a remote data integrity auditing protocol appeared, making cloud users verify the integrity and correctness of the original data without accessing to the original data, and is especially suitable for the cloud computing environment. Therefore, the research to the integrity verification auditing protocol of the cloud data has important practical significance.This thesis focuses on the study of the data integrity verification in cloud storage environments. We analyze the existing data integrity verification schemes by using basic theoretical knowledge in cryptography. Moreover, we elaborate the safety and efficiency of some of the protocols. Our main contributions can be summarized as follows:1. This thesis analyzes the Wang et al.’s auditing protocol, and points out its serious secure flaws. It can not withstand external attacks and malicious cloud servers replaying response values attacks. For the vulnerabilities of the protocol, we give the specific attack scenario.2. Based on existing auditing protocol, we propose a new auditing protocol which is based on Reed-Solomon code and suitable for distributed cloud storage environments. The use of random mask technology ensures data security. The proposed scheme can not only verify the completeness of the data, but also locate the position of the error cloud server when errors are found. Taking advantage of the index table, it achieves complete dynamic operations. In addition, this thesis also analyzes the performance of the protocol with existing ones, and the results shows that our protocol can safely and efficiently achieve the integrity verification for the user’s data in distributed storage scenario.3. Research on Zhu et.al’s multiple cloud data storage auditing protocol and its attack scheme proposed by Wang et.al. Then we analyze the causes of vulnerabilities. In order to avoid these security vulnerabilities, based on the verification framework for multicloud, we proposed a new auditing scheme using elliptic curve discrete logarithm problem. By using the homomorphic verifiable and the public key technology, it reduces communication complexity. The utilization of hash index hierarchy makes it suitable for distributed storage. Using random mask technology and zero- knowledge proof theory, it implements the privacy-preserving problem and storage safety of the data.