Regenerating-Codes-Based Efficient Remote Data Checking and Repairing in Cloud Storage

Abstract

The dramatic development of cloud storage services has led growing companies and individuals to outsource their data to cloud. However, users still concern about the availability and integrity of the data stored in cloud. To relieve these concerns, data redundancy is introduced into cloud storage systems, and data integrity verification schemes are used to check whether data is corrupted. Once data corruption is detected, the repair operations should be executed. However, most of the existing schemes based on erasure codes or network coding techniques either introduce high computation cost or cannot efficiently support remote data repairing. In this paper, we propose an efficient Remote Data Checking and Repairing (RDCR) scheme based on the minimum bandwidth regenerating codes. Our scheme reduces data owners' burden of checking data integrity by enabling a third party to perform the public integrity verification. In addition, unlike previous schemes, our scheme supports exact repair of corrupted data so that the computation cost is further reduced. We implement our scheme and the experiment results show that, compared with the existing schemes, RDCR has lower computational overhead and communication cost.