摘要
针对云雾融合物联网系统中感知数据的质量差异和隐私泄露问题,提出一种基于流加密模式的隐私保护真值发现机制。首先,云服务器将洗牌算法和流加密算法相结合,实现匿名更新真值;雾服务器匿名更新权重,防止恶意攻击者与云雾服务器的共谋攻击所造成的设备隐私泄露。其次,雾服务器基于Softmax函数计算设备权重,以减小计算真值的误差。最后,理论分析证明该机制能够保护设备隐私。实验结果表明,该机制误差率低,并在时间开销上优于已有方法,是一种适用于规模化物联网高效的真值发现机制。
In order to solve diverse quality and privacy leakage of perceived data in fog-cloud integrated internet of things(IoT),a streaming encryption-based privacy-preserving truth discovery mechanism for IoT was proposed.Firstly,by utilizing shuffling and streaming encryption algorithms,the ground truths and the weights were anonymously updated on the cloud server and fog server,respectively,so that the collusion attacks between malicious attackers and cloud or fog servers could be resisted to defend against privacy leakage of IoT devices.Secondly,by adopting the Softmax function,the device weights were calculated on fog server,which reduces the error rate for calculating the ground truths.Finally,the theoretical analysis proved that the mechanism could protect privacy of the devices.And,the experimental results demonstrate that the proposed mechanism is an effective privacy-preserving trust discovery mechanism for large-scale IoT devices,which can outperform existing ones in computing efficiency.
作者
何英杰
李启伟
孙涵
郜迪
董剑峰
杨书华
HE Yingjie;LI Qiwei;SUN Han;GAO Di;DONG Jianfeng;YANG Shuhua(The 22nd Research Institute of China Electronics Technology Group Corporation,Xinxiang 453000,China)
出处
《电信科学》
2021年第5期124-132,共9页
Telecommunications Science
关键词
物联网
轻量级
隐私保护
真值发现
IoT
lightweight
privacy-preserving
truth discovery
