Privacy-Preserving Trust-Aware Group-Based Framework in Mobile Crowdsensing

In practical mobile crowdsensing (MCS) systems, many cooperative sensing tasks require a group of reliable participants to perform collaboratively. In this article, we address the problem of group formulation in MCS, which aims to recruit highly trusted participants and form a high-reputation group. We propose a novel Privacy-preserving Trust-Aware Group Formation (PTAGF) framework that ensures trust and privacy between the group members. This framework consists of three mechanisms; the member trust assessment mechanism, the group forming mechanism, and the two-layer privacy-preserving mechanism. Furthermore, we prove that the group forming problem is NP-hard, and thus propose a heuristic-based Trust-Aware Group Formulation (TAGF) algorithm. A theoretical analysis is provided, which demonstrates that the proposed framework achieves privacy and security. Finally, we experimentally evaluate the performance of PTAGF on a real-world dataset against two state-of-the-art approaches. The results demonstrate that PTAGF outperforms these approaches in terms of trustworthiness in group selection. Moreover, it achieves reasonable task coverage and running time with different communities size, group sizes, and task scales.