移动群智感知中基于纳什讨价还价博弈的多任务分配策略  

Multi-task allocation scheme based on Nash bargaining model in mobile crowdsensing network

作　　者：李丽[1] 钟晓雄 Li Li;Zhong Xiaoxiong(School of Computer Science&Software,Shenzhen Institute&Information Technology,Shenzhen Guangdong 518172,China;Dept.of New Networks,Peng Cheng Laboratory(PCL),Shenzhen Guangdong 518055,China)

机构地区：[1]深圳信息职业技术学院计算机与软件学院,广东深圳518172 [2]鹏城实验室新型网络研究部,广东深圳518055

出　　处：《计算机应用研究》2025年第4期1191-1197,共7页Application Research of Computers

基　　金：深圳市科技计划资助项目(JCYJ20220530143811027);广东省特支计划资助项目(2021TQ06X117);国家自然科学基金资助项目(61802220)。

摘　　要：在移动群智感知网络多个并发任务的情况下,如何根据用户资源以及任务的质量需求实现多任务的有效分配问题,利用纳什议价解模型在资源分配方面的优势,提出了一种基于纳什议价解的多任务分配策略。该策略将多个用户对多个任务根据不同目标以及质量需求的选择问题映射为一个多方纳什议价博弈模型,并采用空间距离的方法有效求得此多方纳什议价博弈的最优解。感知平台根据此解对多个任务进行统一分配,将每个任务分配给最适合的用户去执行。该策略可以实现用户整体得益的最大化,在保障数据质量的同时,减少同一任务执行用户的数目,有效降低感知平台的激励成本。实验结果表明,所提策略比现有任务分配策略具有更好的整体效用与任务质量满意度。In order to realize the effective multi-task allocation based on user resources and task quality requirements for multiple concurrent tasks of mobile crowdsensing network,this paper exploited the advantage of Nash bargaining model in resource allocation,and proposed a multi-task assignment strategy based on Nash bargaining solution.This strategy mapped multiple users’selection problems of different heterogeneous tasks according to different objectives and quality requirements into a multi-player Nash bargaining game model,and used the space distance method to effectively obtain the optimal solution of this multi-player Nash bargaining game.Based on this solution,the sensing platform uniformly allocated multiple heterogeneous tasks,and assigned each task to the most suitable user.The strategy can maximize the overall profit of the user,reduce the number of users executing the same task while ensuring the quality of the data,and effectively reduce the incentive cost of the sensing platform.The experimental results show that the proposed strategy has better performances than the existing task allocation strategies in terms of overall utility and task quality satisfaction.

关 键 词：移动群智感知 纳什讨价还价模型 任务质量 任务分配 空间距离 

分 类 号：TP393.01[自动化与计算机技术—计算机应用技术]