Secure Data Deduplication in Cloud Environments

Fatema Rashid

doi:10.32920/ryerson.14654145

Abstract

With the tremendous growth of available digital data, the use of Cloud Service Providers (CSPs) are gaining more popularity, since these types of services promise to provide convenient and efficient storage services to end-users by taking advantage of a new set of benefits and savings offered by cloud technologies in terms of computational, storage, bandwidth, and transmission costs. In order to achieve savings in storage, CSPs often employ data dedplication techniques to eliminate duplicated data. However, benefits gained through these techniques have to balanced against users' privacy concerns, as these techniques typically require full access to data. In this thesis, we propose solutions for different data types (text, image and video) for secure data deduplication in cloud environments. Our schemes allow users to upload their data in a secure and efficient manner such that neither a semi-honest CSP nor a malicious user can access or compromise the security of the data. We use different image and video processing techniques, such as data compression, in order to further improve the efficiency of our proposed schemes. The security of the deduplication schemes is provided by applying suitable encryption schemes and error correcting codes. Moreover, we propose proof of storage protocols including Proof of Retrievability (POR) and Proof of Ownership (POW) so that users of cloud storage services are able to ensure that their data has been saved in the cloud without tampering or manipulation. Experimental results are provided to validate the effectiveness of the proposed schemes.

Highlights

In the recent years, data deduplication has been advocated as a promising and effective technique to save the digital space by removing the duplicated data from the data centers or clouds
A major issue hindering the acceptance of cloud storage services by users is the data privacy issue associated with the cloud paradigm
Data is outsourced in its encrypted form, there is no guarantee of data privacy when an honest but curious Cloud Service Providers (CSPs) handles the management of confidential data while these data reside in the cloud

Summary

Introduction

Data deduplication has been advocated as a promising and effective technique to save the digital space by removing the duplicated data from the data centers or clouds. There has been a lot of recent interest in providing adequate user privacy in the cloud [27] and improving existing deduplication techniques [17], there is currently no work that addresses both user privacy need and efficient data deduplication in an integrated fashion” In this Chapter, a novel secure data deduplication framework framework for cloud environment is proposed, which can ensure the privacy of the data in the cloud storage and can allow the CSP to perform data deduplication without compromising the data privacy and integrity. A secure video deduplication scheme in cloud storage environments using the H.264 compression algorithm is proposed assuming that the CSP is semi-honest. The partial convergent encryption scheme is meant to ensure that the proposed scheme is secured against the semi-honest CSP; the unique signature generation scheme is meant to allow for a classification of the encrypted compressed video data in such a way that the deduplication process is further efficiently performed on them. It should be noted that the information used in the POW scheme is the one already generated for the POR scheme

Objectives

Results

Conclusion