Research On The Incentive Mechanism Of Mobile Group Intelligence Perception System Based On Auction And Reputation

With the rapid development of the Internet,the Mobile Crowd Sensing system is increasingly used in various fields,such as environmental monitoring and traffic conditions.Mobile Crowd Sensing system can solve a wide range of technical problems in monitoring,transportation,news,etc.through the masses and professional technicians in the society.However,we did not use the huge base of people to exert more effective power,which made many problems in life unable to be solved more efficiently.Due to the problem of insufficient user participation and too little amount of uploaded perception data,the research on incentive mechanism is particularly important in the Mobile Crowd Sensing system.The difficulty in designing an incentive mechanism is that it is impossible to balance the reasonable distribution of tasks and the rewards paid to the participants.These tasks are extremely chalenging for the payment costs of the platform.Most of the existing incentive mechanisms only focus on whether a given task is to be completed,and rarely consider the user's mobility issues,because not all users will meet the conditions to complete the task at a later time.Participants will not always wait for the system to push the task request,and the location and state of the participant may not be suitable for completing this task at the next moment.In addition to this method of using the payment,there is not much research on the incentive mechanism in other aspects.The current social relationship network in society is extremely large.How to make good use of this aspect also has some research space.Because of the above problems and chalenges in the process of mobile group intelligence perception,this article studies from the following two aspects:(1)Incentive mechanism of the reverse auction based on prediction: Most of the current research focuses on deterministic tasks,and rarely considers the user's uncertain mobility issues.As time passes,the user's position changes,and the task matching degree and cost must be recalculated.So this article adopts a form of reverse auction.The platform publishes a set of tasks,the worker's bid,and the winning bidder completes the task.In particular,this paper proposes a prediction mechanism,because,in this system,most users in real life are in a state of constant movement.The task request sent by the user at this moment may become inappropriate at the next moment because his position has changed,it may deviate from the direction of the task and cost more.This paper uses a semi-Markov model to predict the user's location at the next moment.According to the proposed winner selection mechanism,by assigning each task to the previously predicted workers who are most likely to complete the task,then this paper proposes a reward determination mechanism by considering each winner 's moving cost and bid price Decide on appropriate payments to minimize the social cost from all winners.Extensive experiments on real data sets prove that this mechanism has a remarkable performance.(2)Incentive mechanism based on reputation value: To obtain real,reliable,and effective perception data,it is key to formulate a reasonable mechanism to encourage a large number of users to participate.This article introduces social relations into the mobile group intelligence perception system,using the user's position in society to achieve the effect of motivation.Aiming at the selfish problems of participants in mobile group intelligence perception,such as considering their privacy,falsely reporting data or being lazy to obtain maximum compensation,and considering the loss of their mobile devices,etc.,an incentive mechanism based on the reputation value of nodes is proposed.First of al,this paper designs a set of reasonable calculation methods for calculating and updating the reputation value.Secondly,an incentive model is established.The incentive model includes the selection of the user module and the reward and reward module and sets several factors to decide to provide to the participant's compensation.Then according to the platform to score the satisfaction of the participants to complete the task,update their reputation value to complete the subsequent tasks.Finally,the experimental results show that the incentive mechanism proposed in this paper has better effects and feasibility than traditional reputation incentives.