A Framework for Data Integrity Checking In Cloud Computing

Abstract

Cloud computing has emerged as a utility computing paradigm that offers great potential for storing data remotely. It efficiently reduces data maintenance pressures of an organization. However, preserving data integrity and resolving security concerns of outsourced data still remains to be a challenge for data owners, due to loss of control over data. Thus, integrity checking of outsourced data in cloud computing environments becomes crucial for identifying loss of data or any corruption incident in timely manner. Data integrity checking significantly reduces the impact of risks caused due to such losses.

In recent years, because of the growing trends of outsourcing data, significant progress has been made by researchers for designing efficient data integrity checking schemes. Most of the existing solutions have not been able to achieve the desired requirements of integrity checking schemes like public auditing of dynamic data, unlimited times of integrity checking and providing data freshness guarantees etc. Moreover, less attention had been given to minimize computation overhead incurred in integrity verification. Hence, new data integrity checking schemes for cloud computing are needed that fulfill the desired requirements and are computationally efficient

Addressing the issue of integrity checking of outsourced data through literature survey, a comparative analysis of existing data integrity checking schemes is presented in the thesis. Detailed review of cryptographic and mathematical methods used to devise data integrity checking schemes has been carried out. A specific analysis based on security methods, storage overhead, computation cost and communication cost is presented in tabular form. Further, a comprehensive view of desirable properties of these schemes is presented. The challenges in the composition of an efficient data integrity checking scheme for cloud computing environment have been highlighted.

To achieve an efficient data integrity checking scheme in cloud computing which will provide support for desired requirements of integrity checking schemes, the techniques of BLS signatures and traditional Merkle Hash Tree (MHT) have been explored. Finally, a computationally efficient Relative Indexed and Time Stamped Merkle Hash Tree (RITS-MHT) based data integrity checking framework for cloud computing has been proposed. In proposed RITS-MHT framework, public auditing of data has been achieved by introducing Third Party Auditor (TPA) between data owner and cloud server. RITS-MHT efficiently reduces computation burden in searching a node in traditional MHT from O(n) to O(log n), where n is number of leaf nodes.

RITS-MHT is applicable to files outsourced at cloud servers. The proposed framework has been implemented using Pairing Based Cryptography (PBC) and Openssl libraries of cryptography. The performance of RITS-MHT has been tested w.r.t metrics like communication cost, computation cost and storage overhead. To test the computational efficiency of RITS-MHT, integrity verification cost has been computed for varying file sizes from 10 KB to 100 KB. A comparison of performance of RITS-MHT with state-of-the-art protocols has also been done. The security of proposed framework has been evaluated in a Random Oracle Model (ROM) by simulating a game between verifier of data and an adversary. This security model is based on the hardness of Computational Diffie Hellman Problem (CDHP). The proposed framework is deliberately designed to support public auditing and data dynamics of outsourced data. The performance and test analysis shows that the proposed framework is computationally efficient and secure.

Based on the experimental results, it can be concluded that BLS signatures and MHT data structure can be used to devise a computationally efficient data integrity checking scheme. This evaluation is utilized as a basis for propositions of some unconsummated ideas that may be realized in future work.