Abstract: In Software as a Service（SaaS）, tenants can customize multiple duplicates and pay for use. However, plain-text data duplicates are vulnerable to the conspired attacks of malicious service provider. It is important to adopt the confusion strategy to enhance duplicates storage security. This paper introduces the Tenant Duplicates Data Obfuscation model（TDDO） based on linear hidden to resist the conspired attack of service provider. TDDO can make duplicates different with each other to prevent untrusted service provider deleting the whole copy of tenant data. Based on Monte Carlo random monotone function, an extended TDDO model with query keyword ordering strategy is promoted to improve the query efficiency of obfuscation duplicates. Experimental results show that the extended TDDO model has better query performance.

Key words: Software as a Service（SaaS）, multi-tenants, data duplicates, data obfuscation, line hidden

摘要： 在软件即服务模式下，针对租户定制的副本数据在明文存储情况下容易被服务提供商合谋欺诈问题，提出基于线性隐藏的数据混淆模型，通过混淆使得存储相同数据的租户副本具有不同的数据内容表现，防止服务提供商为节省存储空间，整个删除租户不常用数据副本，欺骗租户。使用蒙特卡罗随机单调函数对混淆模型进行拓展，制定关键字查询保序策略，可以在混淆后的租户副本数据上进行查询操作。实验结果表明，扩展的混淆模型在保序关键字上的查询性能接近明文查询。

关键词: 软件即服务（SaaS）, 多租户, 数据副本, 数据混淆, 线性隐藏