Research On Incentive Mechanism For Spatiotemporal Coverage In Vehicle Crowd Sensing

Vehicle crowd sensing is a kind of sensing mode that uses the vehicle with pre installed sensors to obtain data.The applications of vehicle crowd sensing include air quality detection,traffic congestion monitoring,traffic accident reporting,etc.The design purpose of vehicle crowd sensing system is to achieve high space-time coverage with low deployment and maintenance cost.Spatiotemporal sensing coverage refers to the number of spatial grids and time points covered by the collected data.One of the main characteristics of vehicle crowd sensing is that perceptual coverage is very important in both spatial and temporal dimensions.Most of the existing work focuses on encouraging participants to collect data at a specific time and location in order to improve the spatiotemporal perception coverage.However,there are still the following problems: 1)when paying the participants,except for a few,most of them do not consider the impact of the participants’ winning probability on the expected utility.Even if these very few consider the winning probability,they do not consider the overestimation of the probability;2)Only the absolute income of participants is considered,and the influence of reference point on participants’ choice is not considered.However,the prospect theory of behavioral economics shows that in some cases,participants will overestimate the probability of winning,which will affect the expected utility of participants;At the same time,the study of reference point effect found that when participants need to make a decision on an event,they will select a reference point as the reference basis and make a decision according to the relative value rather than the absolute value of the reference point.Based on the above two problems,the main work of this thesis is as follows:First,Payment Mechanism based on the Overestimation of small Probability(PMOP)is proposed.According to the degree of overestimation of small probability,the participants are divided into two stages.In the first stage(cultivation stage),the participants’ overestimation of the winning probability is improved by overestimating the small probability theory.In the second stage(formation stage),participants’ overestimation of small probability is used to reduce the total payment and improve participants’ utility.Secondly,Participant Selection mechanism based on the Reference Point effect(PSRP)is proposed to affect the income reference point of participants by setting the target income,so as to improve the probability of participants accepting tasks.According to the theory of reference point effect,the platform selects participants with high probability of accepting tasks,and improves the space-time coverage while improving the effectiveness of participants.Theoretical analysis and simulation results show that PMOP and PSRP can improve space-time coverage,participant utility and reduce unit task cost.