Research On MCS Participant Selection Based On User-characteristic Aware

Mobile Crowd Sensing(MCS)uses the mobile terminal equipment carried by users as a basic sensing unit,and relies on many users to form a large-scale sensing network,thus completing some complex sensing tasks that are difficult to complete by individuals alone.However,when the service platform selects participants,it is usually limited by budget and time but it has high requirements for data quality.In addition,there are differences in the number of users in different regions.The efforts made by different users to complete the sensing task,the ability to perform contracts,and their own preferences are all different,which has a significant impact on task completion rate and data credibility.Therefore,designing effective participant selection strategies,improving task completion rate,task data quality,and service platform-user bilateral satisfaction are urgent.Thesis introduces the research background,basic characteristics and current research hotspots of MCS networks.Then it analyzes the challenges of how to ensure data quality and improve the satisfaction of users and platforms in the MCS network,and introduces the typical participant selection strategy.A data quality-centered participant selection strategy is proposed for the challenges faced.The regional heat is evaluated according to the number of active users in the area,the average user stay time and the regional history-aware task completion status,combined with the user's status information and user history sensing tasks records calculate the user's willingness,credibility and activity.Comprehensively consider the above four factors to reasonably select the task participants.The results show that the proposed strategy can significantly improve the overall data quality,and can also complete the sensing task in a low-heat region in a timely and reliable manner.A participant selection strategy centered on bilateral satisfaction is proposed.The Mobile Edge Computing(MEC)server evaluates the user's compliance with the sensing task based on the user's real-time status information,and the cloud service platform evaluates the reputation by the historical task records,and then the MEC server establishes the participant revenue perception model by combining the user historical reputation,the participant compliance and the user sensing cost coefficient.At the same time,the task price mechanism is formulated based on the demand index of the data and the supply index of the user.Finally,the participant selection process is transformed into a game between the user and the MEC server regarding price-data quality.Under multiple constraints,the participant with the highest utility ratio is selected,thereby improving the user's profit and satisfaction.The results show that the proposed strategy can significantly reduce the amount of data processed by the cloud service platform,shorten the task completion time,and increase the satisfaction of users and platforms.