Scholarlink

Mobile crowdsensing, while a promising method for handling vast data, encounters two key challenges: accurately estimating the truth of collected data and setting enticing rewards to increase user participation. Most existing studies tackle these issues separately, but our proposed geometry-based data acquisition FrAmework for mobIle cRowdsensing, named FAIR, addresses both simultaneously. FAIR comprises two interlinked parts: FAIR-EST for truth estimation and FAIR-REW for reward determination. Unlike typical methods, FAIR-EST aims to minimize worst-case estimation error, effectively finding the Chebyshev center of the data, which leads to a non-convex quadratic optimization problem solved by semi-definite relaxation. Based on the truth estimated by FAIR-EST, FAIR-REW assigns rewards to maximize platform profit, using an innovative asymmetric all-pay auction model to account for user diversity. Moreover, we offer a rigorous theoretical analysis demonstrating that FAIR can achieve numerous desirable properties under more commonplace conditions.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

FAIR: Accurate Data Acquisition for Mobile Crowdsensing

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • Fuyuan Xia,
  • Chenhao Ying,
  • Wei Chen,
  • Xikun Jiang,
  • Liwei Lin,
  • Yuan Luo

摘要

Mobile crowdsensing, while a promising method for handling vast data, encounters two key challenges: accurately estimating the truth of collected data and setting enticing rewards to increase user participation. Most existing studies tackle these issues separately, but our proposed geometry-based data acquisition FrAmework for mobIle cRowdsensing, named FAIR, addresses both simultaneously. FAIR comprises two interlinked parts: FAIR-EST for truth estimation and FAIR-REW for reward determination. Unlike typical methods, FAIR-EST aims to minimize worst-case estimation error, effectively finding the Chebyshev center of the data, which leads to a non-convex quadratic optimization problem solved by semi-definite relaxation. Based on the truth estimated by FAIR-EST, FAIR-REW assigns rewards to maximize platform profit, using an innovative asymmetric all-pay auction model to account for user diversity. Moreover, we offer a rigorous theoretical analysis demonstrating that FAIR can achieve numerous desirable properties under more commonplace conditions.