Research On Social Attribute Aware Participant Selection Strategy For MCS

Mobile Crowdsensing(MCS)relies on intelligent devices carried by ordinary users as sensing units to collect and upload sensing data,which has become an effective data collection paradigm in the Internet of Things.In a MCS system,how to select effective participants to complete the sensing task is one of the main challenges currently faced by the MCS system.First of all,due to the different real-time status and professional ability of participants,the task completion rate is different,and the quality of submitted data is significantly different.Secondly,different participants have different preferences and different attitudes towards tasks.Participants want to perform tasks they are more interested in.Finally,participants need to pay a certain cost to participate in the sensing task,and the platform needs to provide some incentives to encourage participants to actively participate in the sensing task.Therefore,in order to solve the above-mentioned problems in MCS,this thesis designs the MCS participant selection strategy with social attribute awareness.Firstly,this thesis discusses the MCS research background,research significance and the research status of participant selection.It introduces MCS system knowledge,including MCS network architecture,research hotspots,basic features and typical applications.It analyzes the challenges in MCS participant selection and introduces the relevant theories involved in the participant selection.Secondly,an edge-cloud collaborative MCS participant selection strategy with preference awareness is proposed.Participant preference is quantified in consideration of time factor,distance factor,task type factor,and reward factor.Task preference is quantified in consideration of participant reputation and sensing cost.Based on the preferences of both parties,the participant selection problem is modeled as a stable matching problem between the two parties,and the stable matching is solved to maximize participant preference.A two-stage reward distribution mechanism is proposed to provide appropriate rewards for participants.The results show that the proposed strategy can ensure the quality of task completion and improve participant preference.Thirdly,a multi-stage MCS participant selection strategy with reliability awareness is proposed.The task completion rate and the qualification rate of completed tasks are calculated according to the participant's historical task records.Then the participant reliability is evaluated.This thesis designs a multi-stage participant selection process,calculates the maximum price of the task according to the task requirements at each stage,selects participants who meet the quality requirements for each task,and performs a bargaining game with the participants to obtain the participant's optimal reward;Considering the situation of participants withdrawing midway,a task delegation mechanism is proposed to ensure the effective completion of tasks.The results show that the proposed strategy can guarantee the task completion rate and completion quality,and optimize the interests of participants and the platform.Finally,this thesis summarizes the research content and prospects for the follow-up work.