Incentive Mechanism For Crowd Sensing Based On The Trust Relationship

The mobile crowd sensing(MCS)network has a wide coverage and low node deployment cost.In recent years,MCS has been widely used in various fields,such as environment,public facilities,social networks and social perception.The main feature of this new network is people-centric,and its data collection mode is based on a large number of mobile users.The number of participating users directly determines the success or failure of this MCS application,so it is particularly important to design an efficient and feasible incentive mechanism to improve users' participation rate.The research on incentive mechanism should focus on solving the following problems.First of all,it can improve the users' participation rate,which is the main goal of designing the incentive mechanism,stimulating the enthusiasm of potential users to participate in the task.A large number of users' participation is the basis of the MCS platform.However,a large number of users' participation does not guarantee that the perceived reports they submit are also of high quality.Therefore,incentives need to be able to motivate users to provide higher quality perception reports,improve task completion efficiency.Finally,the design of the group intelligence perception incentive mechanism must consider the privacy protection of users.In order to complete the perception task issued by the system platform,the task report may involve the user's time,space and other private information.This leakage of privacy information will cause harm to users,which may cause the loss of some experienced old users and affect the final task results and completion.Therefore,it is also very important to ensure the privacy of users during the incentive process.The main works of this paper are as follows:(1)Anonymous trust relationship management mechanism.For the user,who wants an anonymous submission-aware task,while platforms prefer to know the identity of the user who submit the data for better data evaluation.The framework proposed in this paper can effectively solve the contradiction between anonymity and trust.The mechanism includes a reasonable reputation reward and punishment system to support the renewal of positive and negative reputation.Users can use blind ID to submit anonymous perception reports,so that their privacy is reasonably protected.At the same time,the system platform can evaluate the submitted report based on the context information when the user is anonymous.The platform manages a reputation database containing user's identity and reputation value.The reputation of the user depends on his or her previous actions,and the user can not control the reputation updating process.In this way,the private information of the recruited users can be fully protected during the whole process of the task,and the users can participate in the task more confidently.At the same time,the users can get reasonable rewards and punishments according to the quality of the perception reports,so as to achieve the goal of motivating users.(2)A user trust relationship model based on reputation value-activity is proposed.In the MCS system platform,the user can choose to become the demand side,also can choose to become the service side.Indirect interaction is inevitable when users transferring data.This kind of interaction can be quantified,and a new concept named as "trust relationship" is proposed in this paper.The trust relationship is related to two indicators,activity level and reputation value.In the past incentive mechanism,the most of work considered the user's reputation value,but rarely considered the user's activity level.A user who has a high reputation but has not participated in a task for a long time is less likely to be trusted.Therefore,this paper takes the activity and reputation values as indicators to consider the relationship of trust and proposes an evaluation plan,this is more fair to the new active users,the system in the retention of old users,but also to stimulate new users.At the same time,it can recruit more trustworthy users,which plays a vital role in improving the accuracy of mobile swarm intelligence system results.(3)Proposed a multi-task user recruitment plan to ensure data quality.In the real scenario,the system platform divides a large task of the demanding user into a series of small subtasks.Each user has a different set of subtasks to sign up for,and the platform wants to recruit users who can provide high-quality,data-aware data.In this paper,we abstract this kind of multi-task user recruitment plan into MTUR problem and prove that the MTUR problem is a NP-hard problem,and at the same time design and propose a recruitment algorithm of MTUR and verifies the performance of the algorithm through the simulation experiment.Under the condition of guaranteeing the completion rate,the users with strong trust relationship are recruited first to guarantee the data quality.The quality and accuracy of perceived reports can be guaranteed by recruiting service users who have strong trust relationship with demandside users.The above solution can effectively let users participate in tasks without worrying about privacy leakage,and stimulate the enthusiasm of potential users to participate in tasks.At the same time,a reasonable reward and punishment mechanism will allow users who provide high-quality perception reports to get more incentive and stimulate users to tend to provide higher-quality perception reports.At the same time,the recruitment plan guarantees the degree of task completion,thereby improving the overall task completion efficiency.