Multi-stage Optimization of Incentive Mechanisms for Mobile Crowd Sensing Based on Top-Trading Cycles

Jingjie Shang, Haifeng Jiang, Chaogang Tang, Huaming Wu, Shuhao Wang, Shoujun Zhang

Abstract: For collaborative tasks requiring multiple users, in Mobile Crowd Sensing (MCS), low user interest in certain tasks usually results in insufficient user re-cruitment. However, the interest of the user directly affects the quality and effi-ciency of task completion. To address this issue, we propose a multi-stage incen-tive mechanism based on the Top-Trading Cycles (TTC) from economics, ena-bling users to participate in tasks that align with their interest through the optimi-zation of multiple stages. Firstly, we perform an initial screening of users using a reverse auction. Then, we adopt the Top-Trading Cycles algorithm to determine the optimal task-user pairs. For tasks with insufficient collaborators, an interest-based task recommendation algorithm is proposed, which calculates user interest similarity in the social network, recommends tasks to other users, and evaluates rewards based on their contributions. The proposed incentive mechanism can theoretically guarantee computational effectiveness, truthfulness, and individual rationality in this paper. Simulation experiments show that the proposed mecha-nism outperforms traditional incentives in terms of user participation rates, task coverage, and average user utility.