Research Of Spatial-temporal User Incentive Mechanism In Crowdsensing

With the process of mobile Internet and the popularization of mobile devices,mobile Internet has penetrated into every aspect of life,completely subverting the traditional mode in the fields of food,clothing,housing and transportation.Crowdsensing is becoming more and more popular,which has great advantages in the scene of collecting,analyzing and processing a lot of data.Due to the constantly changing environment of mobile devices,so how to motivate users effectively based on spatial-temporal requirements will be a challenging problem.At first,there exists dominant and recessive spatial-temporal characteristics of sensing tasks.The user incentive mechanism problem of dominant spatial-temporal correlation is transformed into a set cover problem,which is solved based on the greedy algorithm.Utilizing markov model to predict the spatial-temporal characteristic.Combining the dominant spatial-temporal correlation algorithm and markov prediction model is a way to solve the user incentive mechanism problem of recessive spatial-temporal correlation.Then,the experiments of simulated data and real data sets show that the dominant and recessive spatial-temporal correlation algorithms are effective.Finally,the incentive mechanism based on spatial-temporal correlation ensures the acquisition of high-quality data and the stability and robustness of the whole system.Considering the budget constraint when publishing sensing task,so how to motivate high-quality users in the case of budget constraints becomes our problem.Firstly,we predict the user's spatial-temporal status based on the model of Entropy and Markov.We define the devotion of the user set from the perspective of set cover and spatial-temporal status.We propose the devotion based greedy algorithm to select the greatest devotion user set under budget constraint.Next,the experiments of simulated data show the effectiveness of user incentive greedy algorithm under budget constraint.Finally,budget constraint can reduce the cost of the system,realize the win-win situation with high-quality users,and promote the virtuous cycle of the system.By contrasting the existing works,the thesis creatively study the problem of dominant and recessive spatial-temporal user incentive mechanism,which takes full consideration of budget constraint.The experiments of simulated data and real data sets show that,contracting to other conventional algorithms,the based on spatial-temporal correlation incentive mechanism and the based on spatial-temporal constraints incentive mechanism under budget constraint get better performance in cost,user selection number,iteration number,devotion.Proposed spatial-temporal user incentive mechanism achieve the goal of collaboration and double-win between Crowdsensing platform and users,and ensures the stability of whole Internet,which provides suggestions and ideas for the future research.