Abstract

In the Internet of Things architecture, the distributed management structure of the data center-micro data center (MDC)-intelligent terminals is the foundation for intelligent terminals accessing, real-time data interaction and data release. It is necessary to perform security level detection before intelligent terminals accessing to the MDCs, which can facilitate MDCs to understand the ability to resist malicious attacks and realize reasonable use of terminals in different security levels. However, the dataset of security levels will be stored in MDCs and transferred as terminals removed. Therefore, an expedite privacy protection method for this dataset is required. This paper studies the privacy protection schemes based on differential privacy (DP) protection and proposes a level-proportion-based differential privacy protection method, utilizing the security level and the level proportion of the intelligent terminals as the parameters to apply DP protection with different intensities, so that the statistical properties of the dataset will not be destroyed. Simulation results show that our method can discriminatively implement DP protection for intelligent terminals with different levels. Moreover, it can hold the statistical properties of the dataset for further utilization.

中文翻译：

---

基于差异隐私的智能终端安全等级保护

摘要

在物联网架构中，数据中心-微型数据中心（MDC）-智能终端的分布式管理结构是智能终端访问，实时数据交互和数据发布的基础。在智能终端访问MDC之前，有必要进行安全级别检测，这可以使MDC理解抵御恶意攻击的能力，并实现不同安全级别的终端的合理使用。但是，安全级别数据集将存储在MDC中，并在删除终端后进行传输。因此，需要针对该数据集的快速隐私保护方法。本文研究了基于差分隐私（DP）保护的隐私保护方案，并提出了一种基于等级比例的差分隐私保护方法，以安全级别和智能终端的级别比例为参数，以不同强度应用DP保护，以免破坏数据集的统计特性。仿真结果表明，该方法可以区分级别的智能终端实现DP保护。此外，它可以保留数据集的统计属性以供进一步利用。