Huang Daxin, Gan Qingqing, Wang Xiaoming, Ogiela Marek R, Wang Xu An
Department of Computer Science, Jinan University, Guangzhou, 510632, China.
Department of Cyber Security, Guangdong University of Foreign Studies, Guangzhou 510006, China.
Internet Things (Amst). 2022 Nov;20:100625. doi: 10.1016/j.iot.2022.100625. Epub 2022 Oct 8.
IoT-based crowd-sensing network, which aims to achieve data collection and task allocation to mobile users, become more and more popular in recent years. This data collected by IoT devices may be private and directly transmission of these data maybe incur privacy leakage. With the help of homomorphic encryption (HE), which supports the additive and/or multiplicative operations over the encrypted data, privacy preserving crowd-sensing network is now possible. Until now several such secure data aggregation schemes based on HE have been proposed. In many cases, ciphertext comparison is an important step for further secure data processing. However efficient ciphertext comparison is not supported by most such schemes. In this paper, aiming at enabling ciphertext comparison among multiple users in crowd-sensing network, with Lagrange's interpolation technique we propose comparable homomorphic encryption (CompHE) schemes. We also prove our schemes' security, and the performance analysis show our schemes are practical. We also discuss the applications of our IoT based crowd-sensing network with comparable homomorphic encryption for combatting COVID19, including the first example of privacy preserving close contact determination based on the spatial distance, and the second example of privacy preserving social distance controlling based on the spatial difference of lockdown zones, controlled zones and precautionary zones. From the analysis we see our IoT based crowd-sensing network can be used for contact tracing without worrying about the privacy leakage. Compared with the existing CompHE schemes, our proposals can be collusion resistance or secure in the semi-honest model while the previous schemes cannot achieve this easily. Our schemes only need 4 or 5 modular exponentiation when implementing the most important comparison algorithm, which are better than the existing closely related scheme with advantage of 50% or 37.5%.
基于物联网的群体感知网络旨在实现对移动用户的数据收集和任务分配，近年来越来越受欢迎。物联网设备收集的数据可能是私密的，直接传输这些数据可能会导致隐私泄露。借助同态加密（HE），它支持对加密数据进行加法和/或乘法运算，现在可以实现隐私保护的群体感知网络。到目前为止，已经提出了几种基于同态加密的安全数据聚合方案。在许多情况下，密文比较是进一步进行安全数据处理的重要步骤。然而，大多数此类方案不支持高效的密文比较。本文旨在实现群体感知网络中多个用户之间的密文比较，利用拉格朗日插值技术提出了可比较同态加密（CompHE）方案。我们还证明了我们方案的安全性，性能分析表明我们的方案是实用的。我们还讨论了基于物联网的群体感知网络与可比较同态加密在抗击COVID-19中的应用，包括基于空间距离的隐私保护密切接触确定的第一个例子，以及基于封锁区、管控区和预防区空间差异的隐私保护社交距离控制的第二个例子。从分析中我们可以看出，我们基于物联网的群体感知网络可用于接触者追踪而无需担心隐私泄露。与现有的CompHE方案相比，我们的方案在半诚实模型中可以抗勾结或安全，而以前的方案不容易做到这一点。我们的方案在实现最重要的比较算法时只需要4或5次模幂运算，比现有的密切相关方案有50%或37.5%的优势。
