Research On Location Privacy Protection Based On Secret Splitting

With the rapid development of wireless communication technology and mobile terminals,location-based service(LBS)is becoming more and more popular in people's lives.Location-based navigation,Weibo check-in and We Chat movement are becoming more and more popular in the market,and bring great convenience.However,when enjoying such services,it is necessary to provide the service provider with own specific location,which brings more privacy concerns to users.As a special personal information,location information contains not only location,but also personal information such as identity information,home address and religious belief.Once the information is leaked,it will bring immeasurable loss to personal property and life safety.Therefore,the problem of privacy protection in location service has become increasingly prominent.Most of the existing location privacy protection schemes use k-anonymity,mixed area and differential privacy,but these schemes do not achieve a good balance in privacy and availability.Especially in the case of spatial-temporal correlation between locations in continuous query,most of the schemes are sacrificing availability for privacy.To solve this problem,inspired by Shamir threshold secret sharing scheme,this paper divides the privacy information,so that every service provider in the system can only master a part of the location information,so as to protect the location privacy information.The specific research contents of this paper are as follows:Firstly,a new privacy protection mechanism based on location splitting is proposed to solve the problem of location privacy in proximity detection.Based on the idea of secret split,the mechanism splits location service and location information.On this basis,a dual server split architecture is proposed based on geohash coding.The location geohash code is divided into two parts: location prefix and suffix.Each server only provides services based on partial location information.In addition,a splitting algorithm of prefix and suffix is proposed to balance availability and computing efficiency,which can realize the whole splitting mechanism and protect the location privacy.Finally,compared with other schemes on real data sets,the communication and computation overhead of this scheme is smaller,and the effectiveness of this scheme is verified.Secondly,a sensitive location protection mechanism based on trajectory splitting is proposed to solve the problem of location correlation in continuous LBS.According to the correlation between location on the trajectory,the mechanism proposes the spatial correlation,the travel correlation and the historical correlation.Based on the three location correlations,the privacy risk of sensitive location exposed by publishing non-sensitive location is quantified by differential privacy.By introducing multiple LBS servers to polling scheduling,the non-sensitive location is published to the appropriate LBS server to achieve location "suppression",eliminating the correlation between locations,and thus protecting the sensitive location.Finally,compared with other schemes,this scheme provides better privacy and availability.Finally,aiming at the problem that it is difficult to balance availability and privacy in continuous LBS,a trajectory splitting privacy protection mechanism based on mutual information is proposed.Firstly,from the perspective of information theory,the privacy of trajectory is measured by the mutual information between the disturbed position and the real position on the trajectory,so as to solve the problem that the privacy risk of trajectory is difficult to quantify.On this basis,a theory of trajectory splitting based on mutual information is proposed,and several LBS server architectures are introduced to realize track splitting.The server can achieve scheduling by greedy algorithm to ensure the maximum condition entropy of trajectory.Finally,an optimization algorithm based on Markov model is proposed,which improves the efficiency of the whole scheme.Experiments on real datasets show that the scheme has lower mutual information and better privacy while ensuring the availability,which verifies the effectiveness of the scheme.