A Two-Stage Privacy Preservation Framework for Untrusted Platforms in Mobile Crowdsensing

With the popularization of intelligent terminal devices and the increasing demand for data in the Internet of Things, Mobile Crowdsensing (MCS) has become a new data collection paradigm. At present, user privacy preservation in MCS has attracted great attention, but there are still some shortcomings in the existing research. On the one hand, most research only focuses on privacy preservation mechanisms for either task allocation or data collection independently. On the other hand, most privacy preservation research relies on the fully trusted MCS platform, which is too idealistic in reality. In view of this, we propose a two-stage privacy preservation framework for MCS, which is composed of Differential Privacy based Task Allocation Scheme (DPTAS) and Privacy-aware Heterogeneous Data Collection Scheme (PHDCS). Specifically, DPTAS designs perturbation function based on differential privacy technology to preserve user bid privacy during task allocation. PHDCS designs different data collection methods to ensure user privacy and data quality. Both theoretical derivation and simulation results show that DPTAS can reduce social cost and has excellent performance on privacy preservation. Moreover, the performance of PHDCS is of high accuracy, low time consumption and good capability for privacy preservation.