Task Scheduling And Incentive Mechanism For Mobile Crowd Sensing With Data Sharing

With the rapid development of information science and technology,the Internet-of-Things(Io T)is becoming an extremely important stage in the information age.Io T collects different kinds of useful information through various sensors and forms a huge sensing-enabled network with the Internet,which has been the foundation of Smart City.Nowadays,mobile devices(e.g.,smartphone)have integrated more and more advanced sensors and have powerful capabilities of computing and sensing.The popularity of such mobile devices has as led to the emergence of a new sensing scheme called Mobile Crowd Sensing(MCS).Specifically,MCS creates a new "user-centric" perception model for the Io T,which can achieve a large-scale sensing task scheduling and data collection,and hence has become one of the hot and important research topic.Although there are many research results,there are still many problems in the existing works,such as the duplicated collection of data and false informing by users.Our focus in this work is to study how to solve the duplicated data collection problem,how to achieve the optimal task scheduling with data sharing,and how to motivate users to participate in and meanwhile report their truthful information.In this work,we consider a novel MCS scenario,where different tasks may have common data requirements and can share data with each other.In order to solve the duplicated collection of common data between tasks,we introduce a new data layer between the traditional task layer and user layer,and construct an innovative three-layer network model for MCS.Meanwhile,we present the model with the task,data and user layers in details.Based on such a model,we formulate the optimal scheduling problem(that maximizes the social welfare)as an extended knapsack problem,and solve the optimal scheduling solution by using classic algorithm.Finally,we perform simulations to compare the optimal task scheduling with or without data sharing model which shows the advantages of our proposed three-layer network model with data sharing.Based on the above optimal scheduling scheme,we further stud y the incentive mechanism design in the above model.According to auction theory,we propose a truthful double auction mechanism based on VCG principle.The proposed double auction mechanism includes a scheduling rule and a payment rule,which ensure users and task planners can not get any benefits through false bid,their optimal strategy is reporting the truthful bids.From extensive simulations,we further find that the platform may be in a state of budget deficit.Therefore,we further introduce a reserve price into the double aution and propose an improved auction mechanism which can achieve the platform budget balance.Finally,from simulations,we obtain the relationship between the reserve price,platform budget and social welfare,which can provide guidance for the proper setting of reserve price.In summary,we propose a new model and incentive method from two aspects of task scheduling and incentive mechanism.Besides,we analyze the validity of them by simulation.Our results provide theory supports for the study of data sharing and the related application in MCS.