群智感知系统中面向高斯差分隐私的数据新鲜度性能分析

Gaussian Differential Privacy-oriented Data Freshness Performance Analysis in Mobile Crowdsensing Systems

群智感知是基于众包思想,利用智能感知终端完成传感数据收集的一种数据获取模式,具有部署成本低、实现方式灵活、可扩展性强等优点。随着6G网络技术的日渐成熟,针对基于6G的群智感知系统中亟需解决的传感数据时效性与隐私安全问题,提出了一种基于高斯差分隐私的传感数据内容保护模型,利用信息年龄(Age of Information, AoI)指标对传感数据的新鲜度进行时效性分析,得到了不同队列模型、服务准则以及传输缓存的数据新鲜度性能表达式,突破了传感数据时效性分析与隐私安全提升研究相互独立的现状,为面向隐私保护的群智感知系统时效性性能评估及优化提供理论支撑。通过不同环境参数设置下的仿真实验,所提方案的正确性与有效性得到了验证。结果表明,在典型参数设置下,高斯机制的差分隐私保护效果与传感数据新鲜度性能呈负相关,即高时效性的传感数据隐私安全风险较高,反之亦然。

Mobile crowdsensing is a data acquisition model based on crowdsourcing.By using intelligent sensing terminals to complete sensing data collection,mobile crowdsensing has the advantages of low deployment costs,flexible implementation methods,and scalability.With the development of 6G technology,in order to solve the problem of timeliness and privacy security of sensed data in 6G-based mobile crowdsensing,a content protection model for sensed data based on Gaussian differential privacy is proposed,and the timeliness analysis of the sensed data is performed by using the Age of Information(AoI)metric.Moreover,the mathematical expressions of data freshness are obtained for different queueing models,service rules and transmission cache settings,which can break through the status quo that timeliness analysis and privacy security enhancement research of sensed data are independent of each other.In general,theoretical support for evaluating and optimizing the timeliness performance of privacy-preserving mobile crowdsensing system is provided.The correctness and effectiveness of the proposed scheme are verified through simulation experiments with different environmental parameters.The results show that under typical parameter settings,the differential privacy protection effect of the Gaussian mechanism is negatively correlated with the freshness performance of sensed data,i.e.,sensed data with high timeliness has higher privacy security risks,and vice versa.