A Quality-Validation Task Assignment Mechanism in Mobile Crowdsensing Systems

Participant selection or task allocation is a key issue in Mobile Crowdsensing (MCS) systems. Previous participant assignment approaches mainly focus on selecting a proper subset of users for MCS tasks, but however how to ensure that users devote effort on their tasks is a challenging problem that arises in these approaches. This paper studies the task quality control issue, and proposes a quality-validation task assignment mechanism (QTAM) in MCS systems. We theoretically model this mechanism as a Stackelberg Game, in which the users' instinct of maximizing their payoff and the validation workforce limitation are taken into account. An efficient approximation algorithm is designed to find a Strong Stackelberg Equilibrium (SSE) in the QTAM game. Extensive simulations demonstrate the efficiency and effectiveness of QTAM, which shows that QTAM can prevent untrustworthy behaviors and achieve optimal quality validation for sensing tasks.