车联网群智感知中具有隐私保护的数据聚合方案  

Privacy-preserving Data Aggregation Scheme in Vehicular CrowdSensing

作　　者：李汶锦 蔡英[1] 范艳芳[1] 赵禧玮 于晨 LI Wenjin;CAI Ying;FAN Yanfang;ZHAO Xiwei;YU Chen(College of Computer,Beijing Information Science&Technology University,Beijing 100101,China)

机构地区：[1]北京信息科技大学计算机学院,北京100101

出　　处：《小型微型计算机系统》2025年第1期200-208,共9页Journal of Chinese Computer Systems

基　　金：北京市自然科学基金-海淀原始创新联合基金项目(L192023)资助;国家自然科学基金项目(61973107)资助。

摘　　要：车联网群智感知通过聚合感知数据提高传输效率.由于聚合器不可信,感知数据和聚合结果中所含敏感信息存在隐私泄漏的风险,现有方案采用同态加密或双线性加密实现隐私保护计算和通信开销较大.针对上述问题,本文提出一种具有隐私保护的数据聚合方案,采用基于椭圆曲线的数据掩码技术对数据加密实现隐私保护;采用轻量级的椭圆曲线签名对数据签名以降低计算开销;结合霍纳准则和中国剩余定理对数据进行聚合以降低通信开销.安全性分析证明本方案保证了感知数据和聚合结果的隐私性、完整性、可认证性;性能分析证明本方案与基于双线性映射及基于同态加密的方案相比,在数据收集、数据聚合及结果分析阶段计算开销,端到端通信开销远低于两对比方案.实验结果表明,本文提出的模型在分割任务中具有更好的速度与精度权衡.In Vehicular CrowdSensing(VCS),the transmission efficiency can be improved by aggregating perception data.Because the aggregator is not trusted,there is a risk of privacy leakage of sensitive information such as identity and location contained in the perception data and aggregation results.Existing schemes using homomorphic encryption or bilinear encryption to achieve privacy protection have large computing and communication costs.To address the above issues,a privacy protected data aggregation scheme in VCS is proposed,which uses data mask technology based on elliptic curve non-interactive key negotiation to encrypt perception data and aggregation results to achieve privacy protection;Using lightweight elliptic curve signature algorithms to sign data and aggregate results to reduce computational overhead;Design an aggregation algorithm for perceptual data based on the Horner criterion and the Chinese remainder theorem to reduce communication costs.Security analysis proves that this scheme ensures the privacy,integrity,and authentication of perception data and aggregation results;Performance analysis shows that compared to the MSLPDA(Modified Safe and Lightweight Privacy-preserving Data Aggregation)based on bilinear mapping and the MPPDA(Modified Privacy-preserving Data Aggregation)based on homomorphic encryption,the proposed scheme has significantly lower computational overhead in the data collection,data aggregation,and result analysis stages,and the end-to-end communication overhead is much lower than that of the MSLPDA and MPPDA schemes.

关 键 词：车联网 群智感知 数据聚合 隐私保护 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]