Incentive Mechanism for Mobile Crowdsensing with Two-Stage Stackelberg Game
Chih‐Lin Hu, Kun-Yu Lin, Carl K. Chang
Abstract
Mobile crowdsensing technologies augment the collective effort on exploiting data from a large crowd of mobile users in ubiquitous environments. When mobile users partake in executing crowdsensing tasks, they can receive rewards and be incentified to stay in virtual teamwork. This paper proposes a game-based incentive mechanism, named Incentive-G, aiming at recruiting mobile users effectively and improving the reliability and quality of sensing data against untrusty or malicious users. The Incentive-G mechanism consists of several design phases, including analyzing sensing data, determining reputations of mobile users, and ensuring data quality and reliability by voting in a task group. This mechanism adopts a two-stage Stackelberg game for analyzing reciprocal relationship between service providers and mobile users, and then optimizes incentive benefits using backward induction. Our analysis shows that the existence and uniqueness of the Stackelberg equilibrium can be validated by identifying the best data-provision strategies for mobile users. In addition, the maximum revenue strategy for a service provider can be found by gathering a sufficient amount of high-quality data from mobile users. Performance results manifest that the Incentive-G mechanism is able to significantly encourage mobile users to contribute their efforts and maximize the revenue for game-based crowdsensing services.