Research On Related Technology Of Crowd Sensing Based On Mobile Big Data

With the development of information technology,mobile smart devices grow rapidly,and the era of mobile big data has been quietly arrived.In the era of mobile big data,theory and application of Mobile Crowd Sensing(MCS)has become a hot research fireld both in domestic and foreign academic circles.On the basis of other researchs on MCS incentive mechanisms,for the case users who are limited to their sensory ability and collection ability,a double auction based mobile crowd sensing resource allocation and incentive mechanism was proposed;for the problem that reputation judgement is lacked in online incentive mechanism,a reputaionupdated online incentive mechanisms with budget-limitation was designed;for the two different optimization objectives under the limitation of the number of required sensing tasks,two corresponding incentive mechanisms were put forward.The main research work are summarized as follows:(1)Firstly based on the analysis of the current research status of MCS,research on incentive mechanisms of MCS applications was discussed in detail.Finally,incentive mechanism algorithm model of different objects,different scenes and different targets that both apply auction and game theroy were summarized.(2)An algorithm called DAIM for the resource distribution and incentive mechanism of mobile crowd sensing using double auction was proposed.And using theory and simulation results proved that DAIM satisfied the computationally efficient,individual rationality,budget balance and truthful.At the same time,when taking TASM algorithm which was used in collaborative communication into consideration,DAIM algorithm has higher system utility than TASM algorithm.Thus,DAIM solved the problem that participants who were restricted to their own resources sensing ability,and gave them the solution to ask cloudlets or other smartphone with higher performance for resource distribution request.(3)An evaluation method named ROM(Reputaion-updated Online Mechanism)which combines the user reputation with the data user submitted in online incentive mechanism was proposed.After the theretical analysis and the simulation experiment,it is verified that ROM had better performance than the general online incentive mechanisms,thus guaranteed the data quality that platform allocated,thus guranteed the quality of collected data.(4)Online incentive mechanisms with the limitation of number of all completed task were studied.According to different pratical application,optimization objective was divided into two part,that is,frugal and value maximizing.According to differecnt optimization goals,two mathematic models—frugal task allocating and value maximizing task allocation model were given,meanwhile,time-discounting factor was considered in value maximizing model when value platform could get was calculated.Based on this,two online incentive mechanisms—OZF and OZV were designed,and after the simulation experiment,the performance of these two algorithms were verified.The experiment results showed that both OZF and OZV could achieve better properties than random algorithm.