School of Communication and Information Engineering, Chongqing University of Posts and Telecommunications, Chongqing 400065, China.
Key Laboratory of Optical Communication and Networks, Chongqing 400065, China.
Sensors (Basel). 2018 Nov 15;18(11):3959. doi: 10.3390/s18113959.
Mobile crowdsensing (MCS) is a promising sensing paradigm that leverages diverse embedded sensors in massive mobile devices. One of its main challenges is to effectively select participants to perform multiple sensing tasks, so that sufficient and reliable data is collected to implement various MCS services. Participant selection should consider the limited budget, the different tasks locations, and deadlines. This selection becomes even more challenging when the MCS tries to efficiently accomplish tasks under different heat regions and collect high-credibility data. In this paper, we propose a user characteristics aware participant selection (UCPS) mechanism to improve the credibility of task data in the sparse user region acquired by the platform and to reduce the task failure rate. First, we estimate the regional heat according to the number of active users, average residence time of users and history of regional sensing tasks, and then we divide urban space into high-heat and low-heat regions. Second, the user state information and sensing task records are combined to calculate the willingness, reputation and activity of users. Finally, the above four factors are comprehensively considered to reasonably select the task participants for different heat regions. We also propose task queuing strategies and community assistance strategies to ensure task allocation rates and task completion rates. The evaluation results show that our mechanism can significantly improve the overall data quality and complete sensing tasks of low-heat regions in a timely and reliable manner.
移动众包感知 (MCS) 是一种有前途的感知模式，利用大规模移动设备中的各种嵌入式传感器。其主要挑战之一是有效地选择参与者来执行多个感知任务，以便收集足够和可靠的数据来实现各种 MCS 服务。参与者的选择应考虑到有限的预算、不同任务的位置和截止日期。当 MCS 试图在不同的热点区域有效地完成任务并收集高可信度的数据时，这种选择就变得更加具有挑战性。在本文中，我们提出了一种用户特征感知的参与者选择（UCPS）机制，以提高平台在稀疏用户区域获取的任务数据的可信度，并降低任务失败率。首先，我们根据活跃用户的数量、用户的平均停留时间和历史区域感知任务来估计区域热度，然后将城市空间划分为高热区和低热区。其次，将用户状态信息和感知任务记录结合起来计算用户的意愿、声誉和活跃度。最后，综合考虑这四个因素，为不同的热点区域合理选择任务参与者。我们还提出了任务排队策略和社区协助策略，以确保任务分配率和任务完成率。评估结果表明，我们的机制可以显著提高整体数据质量，并及时可靠地完成低热区的感知任务。
