Data Collection Based On Probability Model And Privacy Protection For Mobile Crowdsensing

The accurate received signal strength(Received Signal Strength,RSS)that an access point can receive is the most important reference for network deployment and planning.While Mobile Crowdsensing provides a new way to build large-scale outdoor RSS maps,it also brings the risk of privacy exposure in user's location privacy disclosure.There are two main problems in the existing Mobile Crowdsensing systems.On the one hand,it increases the risk of leakage of user privacy while adding trusted third parties to further hide users' trajectory information.On the other hand,it is unconsidered the probability of users passing through different road sections is not equal.In this thesis,we address the issues of the compositional structure of group wise sensory systems and the probability that participating users pass through different road sections.How to reduce the groupware intelligence system components and how to design a system based on users with different probabilities to pass road sections to protect users Trajectory privacy algorithm,how to assess the participation of users on all sections of the map coverage point of view of a series of new models and new methods.Experimental results show that the removal of trusted third parties will not reduce the effectiveness of participating users' privacy protection.In addition,the privacy protection algorithm proposed in this paper based on the participants passing through different sections with different probabilities can effectively protect the path privacy of the participating users.The main work of this paper including:(1)Constructing a new Mobile Crowdsensing system.Remove the trusted third party,migrate the function of the third party to the mobile terminal,and reduce the risk of participating in the leakage of user privacy due to the complexity of the system.(2)By analyzing the actual situation that the participating users pass the road sections with different probabilities,constructed a Markov model for predicting the trajectory of the participating users and quantified the level of privacy security of participating users.(3)Give a posterior probability algorithm for calculating the participation of users through the current road segment.This algorithm used to calculate the next link with the highest posterior probability after the participant passes the previous link.