移动群智感知中用户隐私度量与隐私保护研究

Research on User Privacy Measurement and Privacy Protection in Mobile Crowdsensing

针对移动群智感知中,现有隐私保护方案因缺乏有效的隐私度量手段而对所有感知数据采用统一的隐私保护策略,导致对感知数据的隐私信息保护过度或保护不足,且获得的感知数据的准确度较低的问题,文章提出一种面向移动群智感知的用户隐私度量方法,并基于该方法构建一种个性化隐私保护方案。首先,根据感知用户的历史时空数据,利用模糊推理技术,结合该位置所具有的公众属性和该位置对某一用户所具有的不同的个性属性等度量指标进行隐私度量,得出用户在不同位置的隐私度;然后感知平台根据用户上传的不同隐私度,在每个位置选择隐私度较低的用户参与感知任务,以确保用户在隐私安全的前提下,贡献准确度高的感知数据。仿真结果表明,文章方案在提高隐私保护水平的同时确保了感知数据的准确度。

In view of mobile crowdsensing, the existing privacy protection schemes use an unified privacy protection strategy for all perceived data because of the lack of effective privacy measurement methods, which leads to the problem of excessive protection or insufficient protection of the privacy information of the perceived data, and the accuracy of the perceived data is low, this paper proposes an user privacy measurement method for mobile crowdsensing and constructs a personalized privacy protection scheme based on this method. Firstly, according to the historical and spatio-temporal data of the sensing users, by using the fuzzy reasoning technology, combining the public attribute of the locationand the different personality attributes of the location to a user, the privacy measurements of the user in different locations are obtained. Furthermore, the sensing platform selects a sensing user with low privacy measurement to participate in the sensing task in each location according to the different privacy measurement that user uploads, which ensures that users can contribute to the perceived data with high accuracy under the premise of privacy security. The simulation results show that the scheme ensures the accuracy of the perceived data while improves the level of privacy protection.