User Recruitment with Budget Redistribution in Edge-Aided Mobile Crowdsensing

Abstract

Mobile crowdsensing is an efficient data collection method using various mobile smart devices to collect data. Most of the existing mobile crowdsensing frameworks adopt a two-layer architecture model with only a cloud platform to recruit users. However new challenges arise: one is how to reduce the pressure and transmission cost of the cloud platform. Another is how to improve the coverage rate of some areas with low completion rates to ensure data quality. In this paper, we propose an original three-layer mobile crowdsensing framework composed of the cloud platform, edge nodes, and users. It transfers user recruitment and data processing to edge nodes, which offload the data of the cloud platform. Moreover, we propose the offline and online mechanisms based on users’ reputations to solve user recruitment in the edge node. Furthermore, a budget redistribution (BRD) algorithm is proposed. It dynamically redistributes the budget according to the task completion rate of different edge nodes. Finally, we show the proposed mechanism was truthful, individual rationality, calculation efficiency, and budget feasibility. Extensive simulations on real data sets show the reliability and effectiveness of our proposed framework and algorithms.