Abstract
Mobile crowd sensing (MCS) systems usually attract numerous participants with widely varying sensing costs and interest preferences to perform tasks, where accurate task assignment plays an indispensable role and also faces many challenges (e.g., how to simplify the complicated task assignment process and improve matching accuracy between tasks and participants, while guaranteeing submitted data credibility). To overcome these challenges, we propose a service benefit aware multi-task assignment (SBAMA) strategy in this paper. Firstly, service benefits of participants are modeled based on their task difficulty, task history, sensing capacity, and sensing positivity to meet differentiated requirements of various task types. Subsequently, users are then clustered by enhanced fuzzy clustering method. Finally, a gradient descent algorithm is designed to match task types to participants achieving the maximum service benefit. Simulation results verify that the proposed task assignment strategy not only effectively reduces matching complexity but also improves task completion rate.
Highlights
The diversification and popularization of embedded mobile devices enable innumerable user-centric mobile crowd sensing (MCS) applications [1,2,3,4,5,6,7,8,9]
We propose the service benefit aware multi-task assignment (SBAMA) to quickly and accurately match MCS tasks with the most appropriate participants to improve the task completion rate and data credibility
The service benefit of participants is modeled based on their task difficulty, task history, sensing capacity and sensing positivity to improve the accuracy of task assignment
Summary
The diversification and popularization of embedded mobile devices enable innumerable user-centric mobile crowd sensing (MCS) applications (e.g., traffic monitoring, pollution monitoring, and indoor positioning) [1,2,3,4,5,6,7,8,9]. The quality of data uploaded for different task types was not clearly modeled in these studies It may be impossible for the platform to assign tasks reasonably and accurately, resulting in low matching accuracy/task completion rate, high computing resource consumption, and incredibility of data. We propose a service benefit aware (SBAMA) multi-task assignment strategy for MCS. An iterative participant search method based on gradient descent is designed to match participants with the best service benefit in each cluster quickly and accurately. An iterative gradient descent algorithm is proposed to tune the tradeoff between interests of participants and platform It decouples the service benefit from movement distance such that the most appropriate participants for tasks can be found accurately and quickly.
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have