参与者信誉度感知的MCS数据收集机制

在群智感知(mobile crowd sensing,MCS)数据收集过程中,任务参与者的恶意行为能够显著降低感知结果的真实性.为解决此问题,提出了一种参与者信誉度感知的数据收集机制,通过意愿程度和数据质量分析信任状态、量化历史信誉度,进而,根据逻辑回归模型动态更新参与者当前信誉度.同时,为准确衡量感知数据可信程度,利用剩余可发送时间和移动设备剩余能量将参与者分为直接发送和间接转发两类,从而在多任务并发场景下,服务器根据结果合理地选择任务参与者,达到准确可靠收集感知数据的目的.结果表明所提出数据收集机制能大幅度提升感知任务实时性,显著提高感知数据质量,有效降低服务器总奖励开销.

MCS data collection mechanism for participants' reputation awareness

Task participants ' malicious behavior can significantly reduce the credibility of mobile crowd sensing ( MCS ) . To solve this problem, this paper proposed a data collection mechanism that analyzed and quantified participants' historical reputation ac-cording to their willingness and the quality of data they had shared, and then updated their current reputation through the logistic re-gression model. Simultaneously, to measure the authenticity of the collected data, the participants were divided into two types:those who were related to direct transmission of sensing data and second, those who were involved in indirect forwarding of these, which was based on the remaining transmission time of sensing data and residual energy of mobile equipment. Then the server analyzed the accu-racy of data collected by participants according to the multitasking scenario. Simulation results show that the proposed mechanism can significantly improve the perceived tasks performed in real time, greatly upgrade the quality of sensing data, and effectively reduce the reward expenses.