A Trust, Probability Deviation, and Path-Dependence-Based Incentive Mechanism to Improve Data Quality and Coverage in MCS

Mobile crowdsensing (MCS) has emerged as a novel method for decentralized Internet of Things (IoT) environments where collaborative intelligence is fostered through meta-compu-ting. It is urgent to establish an effective incentive mechanism that can enhance task coverage, maximize the data quality, and mini-mize costs for MCS. To address this issue, this paper proposes a Trust, Probability Deviation, and Path Dependence-based Incen-tive (TPD2I) mechanism to improve task quality and coverage at a low cost for MCS. The main contributions are as follows: (1) We introduce a Lightweight Probability Deviation and Path Depend-ence-based Incentive (LPD2I) mechanism, leveraging behavioral economics path dependence theory to promote task coverage while maintaining low costs. LPD2I consists of two stages: the first stage focuses on user training, where users actual earnings exceed the platforms published rewards, encouraging them to take on un-popular tasks, thus enhancing task coverage. The second stage is to maintain high task coverage by utilizing users path dependency and probability estimation deviation, while keeping costs low. (2) A trust-based user identification method is proposed to address the IEWV problem. This method allows the LPD2I mechanism to select trustworthy users as much as possible, which improves data quality. Extensive experimental results demonstrate that TPD2I outperforms existing methods in terms of data quality, task cover-age, and cost efficiency.