The notion of Remote Data Auditing (RDA) has been put forward to achieve the integrity verification for cloud data. However, most of the existing RDA techniques suffer from heavy computational overhead or security attacks. To address this challenge, we introduce an online/offline remote data auditing (OORDA) framework that defines the data auditing process as online and offline phases. Then a concrete OORDA scheme is proposed to ensure secure integrity checking for cloud data. Since some expensive computations are performed offline in advance, the online computational cost can be greatly reduced, which can well solve the performance bottleneck for auditing large-scale data on the auditor. Based on the Computational Diffie–Hellman problem, the proposed OORDA scheme is provably secure in the random oracle model. Performance analysis confirms that the proposed scheme has the optimized efficiency compared with existing schemes, where the average online computational cost can be improved by more than 68% during auditing. Subsequently, a Batch OORDA scheme is presented to support batch data auditing operations, which can reduce the number of pairing operations to constant size at the online phase, independent of the number of data owners. Furthermore, we discuss that the designed file block tag and auditing process can be applied to other related fields, such as verifiable keyword search, verifiable SQL query.

远程数据审计（RDA）的概念被提出来实现对云数据的完整性验证。然而，大多数现有的 RDA 技术都受到大量计算开销或安全攻击的影响。为了应对这一挑战，我们引入了在线/离线远程数据审计 (OORDA) 框架，该框架将数据审计过程定义为在线和离线阶段。然后提出了一个具体的OORDA方案来保证云数据的安全完整性检查。由于提前离线进行了一些昂贵的计算，可以大大降低在线计算成本，很好地解决了审计师对大规模数据进行审计的性能瓶颈。基于计算 Diffie-Hellman 问题，所提出的 OORDA 方案在随机预言机模型中可证明是安全的。性能分析证实，与现有方案相比，所提出的方案具有优化的效率，审计期间平均在线计算成本可以提高68%以上。随后，提出了一种支持批量数据审计操作的 Batch OORDA 方案，该方案可以将在线阶段的配对操作数量减少到恒定大小，而与数据所有者的数量无关。此外，我们讨论了设计的文件块标记和审计过程可以应用于其他相关领域，例如可验证的关键字搜索、可验证的 SQL 查询。提出了一种批量 OORDA 方案来支持批量数据审计操作，它可以将在线阶段的配对操作数量减少到恒定大小，而与数据所有者的数量无关。此外，我们讨论了设计的文件块标记和审计过程可以应用于其他相关领域，例如可验证的关键字搜索、可验证的 SQL 查询。提出了一种批量 OORDA 方案来支持批量数据审计操作，它可以在在线阶段将配对操作的数量减少到恒定大小，而与数据所有者的数量无关。此外，我们讨论了设计的文件块标记和审计过程可以应用于其他相关领域，例如可验证的关键字搜索、可验证的 SQL 查询。