Quality-Enhanced Incentive Mechanism Based On Mobile Crowd Sensing

In recent years,sensor network and mobile Internet develop rapidly.At the same time,the explosive development of smart phones has given citizens the ability to sense and share environment information.Citizens can now easily obtain environment information and share it using mobile Internet.Based on the power of mobile devices,mobile crowd sensing(MCS)emerges as a new sensing paradigm.MCS utilizes the smart phones and wearable devices that distributed everywhere of the city as sensing nodes.The sensing nodes perceived widespread,large-scale deployment in every corner of the city.Since difference level MCS systems have difference levels of user participation,we divide current MCS systems into two categories: systems with active user participation and systems where user participation is passive.Based on the cooperation of the sensing nodes,MSC can perform complex sensing tasks,which makes it a promising research filed and capable of many potential applications.In MCS,we focus on the incentive mechanism in our study.As we know,in MCS,increasing number of average users is allowed to share local knowledge acquired by their smartphones.The employed smartphones to sense will consume their own resources of communication,computation,and energy.It is naturally that users will not participate in the sensing task,unless they are sufficiently motivated.Therefore,a well-designed incentive mechanism is needed for MCS systems.The main work is as follows:1)We propose a novel framework for incentive mechanism based on MCS.The purpose of this framework is to encourage people to participate in the crowd sensing tasks and collect data.It combines the user track information in the physical world and the check-in information with LBSN(Location-based Social Network)in the digital world,with which we can select workers actively and can easily estimate the value of a task.What's more,reverse auction is used to incentive users participate in the sensing tasks.2)With incentive mechanism,we can have enough number of users and large amount of data for sensing tasks in mobile crowd sensing system.However,it cannot serve coverage-oriented MCS tasks.Thus,we propose a novel data selection method based pyramid-tree data clustering approach.The result of data selection is used for multi-payment scheme based on reverse auction.Experiments based on real-world data prove the validity of our method in data coverage and incentive user participation.3)Whether to provide high quality data service is an important factor that affects the usage of mobile crowd sensing system.Thus,we do some study in data quality in this paper.In order to improve the fairness and credibility of the system in the evaluation of data,we make a comprehensive evaluation of the data submitted by the fuzzy logic system from two aspects: subjective and objective.In addition,we propose a reputation model based on time recency,which will be considered for user selection The experimental results show that the model can improve the data quality for the mobile crowd sensing system.