Communications and Networking. 11th EAI International Conference, ChinaCom 2016, Chongqing, China, September 24-26, 2016, Proceedings, Part I

Research Article

Incentive Mechanism for Crowdsensing Platforms Based on Multi-leader Stackelberg Game

Download
153 downloads
  • @INPROCEEDINGS{10.1007/978-3-319-66625-9_14,
        author={Xin Dong and Xing Zhang and Zhenglei Yi and Yiran Peng},
        title={Incentive Mechanism for Crowdsensing Platforms Based on Multi-leader Stackelberg Game},
        proceedings={Communications and Networking. 11th EAI International Conference, ChinaCom 2016, Chongqing, China, September 24-26, 2016, Proceedings, Part I},
        proceedings_a={CHINACOM},
        year={2017},
        month={10},
        keywords={Incentive mechanism Crowdsensing Pricing strategy Stackelberg game Nash equilibrium},
        doi={10.1007/978-3-319-66625-9_14}
    }
    
  • Xin Dong
    Xing Zhang
    Zhenglei Yi
    Yiran Peng
    Year: 2017
    Incentive Mechanism for Crowdsensing Platforms Based on Multi-leader Stackelberg Game
    CHINACOM
    Springer
    DOI: 10.1007/978-3-319-66625-9_14
Xin Dong1, Xing Zhang1,*, Zhenglei Yi1, Yiran Peng1
  • 1: Beijing University of Posts and Telecommunications
*Contact email: hszhang@bupt.edu.cn

Abstract

Nowadays, the exponential growth of smartphones creates a potential paradigm of mobile crowdsensing. A sensing task originator accomplishes its sensing data collection work by publishing them on crowdsensing platforms. All the platforms want to attract the task originator to use their services in order to make higher profit. Thus, the issue of competition arises. In this paper, we study the incentive mechanism based on pricing strategy for crowdsensing platforms. We formulate the competition among platforms as a dynamic non-cooperative game and use a multi-leader Stackelberg game model, where platforms are leaders and the task originator is the follower. In the real world, it is difficult for a platform to know the strategies of others. So we propose an iterative learning algorithm to compute its Nash equilibrium. The iterative learning algorithm is that each platform learns from its historic strategy and the originator’s response. Through extensive simulations, we evaluate the performance of our incentive mechanism.