A rapid detection algorithm of corrupted data in cloud storage

- A rapidly detecting algorithm based on three-dimensional data locating is proposed.
- The consistent hash is utilized to balance the computation cost.
- The cube splitting is applied to locate the corrupted blocks.
- Data concatenated and data blind technology are utilized in detection process.

The cloud computing provides dynamically scalable and virtualized resource service for users to access the storage data. Although having been bringing enormous convenience, it also incurs the threat of users' data loss or corruption, such as data intentionally deleted or corrupted, the service providers' hardware error and careless operation. Most of the data verification schemes based POR or PDP are proposed to verify the integrity of a data block or even a batch of data blocks. However, once the batch verification fails, it results in that all the blocks in the batch of data cannot be judged to be intact or corrupted since the corrupted blocks are not accurately identified. To improve the efficiency of corrupted data identified, we propose a rapid detection algorithm of corrupted data based on three-dimensional data locating, which is called cube-based detection. Furthermore, the consistent hash is utilized to balance the computation cost, and the cube splitting is applied to locate the corrupted blocks by narrowing the range of suspicious blocks step by step. Finally, both data hierarchically concatenated and data blind technology are utilized to improve the verification efficiency and preserve users' data privacy in the detection process. Theoretic analysis and simulation results demonstrate that our algorithm has strong detection capability and identification capability for all the corrupted blocks, and greatly decreases the cost of verification data transmission and computation.