Joint Location-Value Privacy Protection for Spatiotemporal Data Collection via Mobile Crowdsensing

Due to the development of the Internet of Things, mobile crowdsensing has emerged as a promising pervasive sensing paradigm for online spatiotemporal data collection, by leveraging ubiquitous mobile devices. However, privacy leakage of device users is a crucial problem, especially when an untrusted central platform in mobile crowdsensing is considered. Moreover, private information of users like trajectories contained in both location tags and sensed values of their sensing data may be unexpectedly revealed to the platform. In order to solve this problem, we proposed a joint location-value privacy protection approach, which consists of two privacy preserving mechanisms to perturb the locations and sensed values of users, respectively. The approach can be performed by each user locally and independently. The privacy of users can be well preserved, as we theoretically prove that the two mechanisms satisfy local differential privacy. In addition, extensive simulations are conducted, and the results show that accurate estimated values can be derived based on perturbed locations and sanitized sensed values, by adopting the truth discovery method.