Research On Reverse Auction Incentive Mechanism For Crowd Sensing

With the surge of mobile electronic devices and the development of smart cities,crowd sensing has emerged as a new way of sensing data.Unlike previous wireless sensor networks that need to be deployed in advance,the data collection work of crowd sensing relies on mobile users of smart devices with sensing functions.Which means the server platform mainly relies on the "crowd" in the process of sensing data collection.The human group is not only numerous,but also uncertain.Therefore,how to design a reasonable incentive measure to attract users to participate in activities is the main challenge currently.Taking into account the constraints of objective factors such as selfish psychology and time and space that users are unwilling to provide free perception services,based on the game theory method,two different reverse auction incentive mechanisms,offline and online,are proposed.(1)In view of the fact that most of the existing incentive mechanisms focus on maximizing platform benefits and users are generally not enthusiastic about participating,an offline incentive mechanism centered on reverse auctions is proposed.In the bidding stage,comprehensive selection is made according to the task value and quotation in the user's hands;considering that most of the existing payment methods are paid in a uniform amount,which is not conducive to distinguishing contributions,In the payment phase,the first place in the set of winning bidders is given an additional payment method.Finally,through simulation experiments,platform using utility,average user utility and task coverage as indicators,it is finally verified that the mechanism has a good incentive effect in terms of efficiency,individual rationality,feasibility and authenticity.(2)Aiming at the characteristics of user mobility and randomness,the existing offline incentive mechanism cannot meet the wider online application scenarios,and an online incentive mechanism centered on reverse auctions is proposed.Considering that most of the existing online incentive mechanisms only consider a single variable(such as user quotations,the number of tasks completed,etc.),this dissertation comprehensively considers time,data quality,and quotation factors to assess user credibility.In the bidding stage of users who choose to perform tasks,a phased and sub-group approach is adopted.At each stage,the highest credibility is selected to win;considering the long-term participation of users,a certain reward is given to the unsuccessful according to the value of each user's contribution.This allows users who fail in the previous stage to continue participating in the activity.Finally,through simulation experiments,it is verified that the incentive mechanism has achieved better incentive effects from the aspects of active user participation,task value and user utility.Figure[18]table[3]reference[75]...