Achieving Privacy-Preserving Multi-Task Allocation for Mobile Crowdsensing

Yuanyuan Zhang,Zuobin Ying,C. L. Philip Chen

DOI: https://doi.org/10.1109/jiot.2022.3153473

IF: 10.6

2022-01-01

IEEE Internet of Things Journal

Abstract:In the mobile crowdsensing (MCS) with large-scale data collection and sharing environments, since a growing number of applications need to exploit multisource sensing information, it is almost indispensable to develop a generic mechanism supporting efficient and accurate multiple tasks allocation. Meanwhile, achieving the maximum service benefit, the cloud server allocates the multitask based on the user attribute preferences, but it will lead to the privacy leakage of sensing users (SUs). Motivated by the aforementioned challenges, we propose a privacy-preserving multitask allocation (PMTA) scheme for MCS in this article. Specifically, we exploit $K$ -means clustering and matrix multiplication to realize a secure and efficient grouping mechanism, which achieves the selection of high-quality and accurate target users set with privacy preserving. Based on the short group signature algorithm and 0–1 encoding technique, we construct a privacy-preserving matching mechanism to guarantee the anonymous authentication and achieve the matching for task requirements and user reputation levels in a privacy-preserving way. Finally, we give a security analysis, and we evaluate the computational costs and communication overhead, and the experimental result shows the efficiency of our proposed PMTA scheme.