Research On Mobile Crowd Sensing For Perceptual Quality Assurance

Mobile Crowd Sensing(MCS)network takes the intelligent device carried by ordinary users as the basic perception unit,and realizes the distribution of perception tasks,collection and upload of perception data through conscious or unconscious collaboration through the network,so as to complete huge and complex social perception tasks.It not only provides a new perception mode for the Internet of things,but also brings a series of new challenges.As a new research field,mobile group intelligence perception network still faces many challenges that traditional sensor networks have not faced.Task allocation is the core problem of mobile crowd sensing data collection.Efficient task allocation methods can optimize the system while completing tasks.The subjective intention of users will largely affect the positive degree of users' participation in perceptual tasks.Only by using reasonable incentive scheme can sufficient users participate in perceptual tasks.Based on this,this paper has carried out research on the task allocation mechanism and incentive mechanism in the mobile group intelligence perception network.The main work of this paper is as follows:Aiming at the situation of perceiving the resource shortage of platform users,a task allocation method for multi-task concurrency is proposed.Different from other task assignment methods,in the case of minimizing the number of participants,the participants are no longer limited to only one perception task,and the perception platform hopes that each participant can complete multiple perceptions as much as possible within the specified time.At the same time,the incentive cost of the sensing platform is reduced by minimizing the total distance moved by the participants.Minimizing the incentive cost and the number of participants as the optimization goal of the task assignment model,On the problem of multi-objective optimization,the model is solved by the combination of Pareto solution set and particle swarm optimization algorithm,and the necessary improvement of the algorithm is proposed for the mutuality requirement.Aiming at the problem of insufficient number of participants and low data quality,a mobile swarm intelligence perceptive incentive model oriented to task cost difference was proposed.Firstly,the task is classified by the task cost.Secondly,the perceived data quality is evaluated by the timeliness,integrity and accuracy of the perceived data,and the perceived quality and time decay factor are used to update the credibility.Finally,the participant's reputation and The task level is perceived by the task level,and the participant realizes the update of the credibility and the reward of the reward by completing the perceptual task and uploading the perceptual data.Through the interaction between the modules,the purpose of effectively recruiting more users and improving the quality of perceived data is achieved.