Privacy Protection Based Incentive Mechanism For Mobile Crowd Sensing

As a new type of participatory sensing pattern,Mobile Crowd Sensing(MCS)takes advantage of mobile intelligent devices to collect data for large-scale and complex sensing tasks.In the process of participation,mobile users inevitably need to consume the resources of mobile devices.Hence,rational mobile users tend not to participate in the tasks gratuitously.Additionally,sensing data may also directly or indirectly reveal users' privacy information,reducing users' enthusiasm to participate.Therefore,the combination of incentive mechanism and privacy protection is very important for the popularization of MCS.The main work of this thesis includes the following three aspects:(1)Based on the analysis of competition and cooperation relationship between mobile users and task platform,the MCS sensing model is defined and a privacy-protection based incentive system framework is designed.The framework introduces a trusted third-party,which plays an important role:as an arbitration institution,it ensures the fairness of the bidding process;as a credit center,it ensures the effectiveness of incentive mechanism;as an accountability center,it can hold the malicious users accountable combined with the task platform;as an agent,it can eliminate the relationship between mobile users and data.(2)From the aspect of privacy protection,combining data type and attack sources,this thesis presents a data-association based threat model,and designs the user-anonymous workflow of sensing task.The security function analysis shows that the combination of third-party and some data encryption technologies(such as blind signatures)can reduce the relevance between users and data,decrease the number of interaction between users and task platform,and thus improve the level of user privacy and security.(3)From the aspect of user incentive,this thesis proposes a credit based incentive mechanism.This mechanism can be divided into two stages.In the user qualification evaluation stage,the subjective and objective factors which affect the data quality are fully considered.In the dynamic user selection stage,the definition and calculation method of data quality relevance,user credibility,user location relevance and user utility are designed,and the contradiction between user utility and bidding price is weighed.A dynamic filtering algorithm based on credit is proposed to optimize the platform benefit,and the calculation method of user compensation is designed.Based on large-scale real data sets,a series of simulation experiments are carried out to verify the effectiveness of the proposed solution.The experimental results show that the incentive mechanism proposed in this thesis has high time efficiency and stability.It can be flexibly adapted to different sensing tasks such as task budget,expected quality and data distribution requirement.In particular,the incentive mechanism proposed can effectively reduce the proportion of low-credit users in the system.