Research On Privacy Preserving Schemes Of Location-based Service In Mobile Internet

With the improvement of intelligent handheld devices,the development of mobile communication technology and the maturity of positioning technology,people can freely access the mobile internet and enjoy the convenience of location-based service(LBS).The application of LBS has played an improtant role in the fields of life services,social communication and entertainment,and has become an indispensable part of people's daily life.However,the queries of LBS contain a large amount of users' sensitive information.If the protection for users'sensitive information is not enough,potential attackers such as location-based service providers can easily obtain users' privacy information via the collection of users'sensitive information.Therefore,how to protect users'privacy has become one of the key issues for the development of LBS,and the research on LBS privacy protection has become an urgent matter.The core and goal of this paper is to guarantee the security of users'privacy while providing high-quality service.Based on the different application fields of LBS,the issues of privacy protection and utility in points of interest(POI)mode,sociality mode and review mode are studied.The main contents of this paper are as follows:(1)In non-outsourcing cloud scenario of POI mode,a POI classification table based privacy preserving scheme is proposed to solve the problems of privacy protection and utility in the architecture of mobile user ad hoc network.Specifically,considering the vulnerability of users'preference privacy,the POI classification table is designed to protect users'preference privacy in the process of ad hoc networking.Based on the premise that users are not trusted in ad hoc networks,the method of additive homomorphism is used to calculate the kinds of POI.This method enables the anchor to calculate the kinds of POI without knowing other users'POI types in an ad hoc network,which can effectively protect users'preference privacy in the process of ad hoc networking.By studying the dynamic relationships between the number of users' locations k and the kinds of POI l as well as the corresponding privacy preserving thresholds(thk,thl,)in the process of ad hoc networking,an adaptive algorithm is designed to improve the efficiency and success rate of ad hoc networks.Theoretical analysis and experiments show that the scheme can effectively protect users' preference privacy in the process of ad hoc networking,and improve the efficiency and success rate of ad hoc networks.(2)In outsourcing cloud scenario of POI mode,a dual clouds based privacy preserving scheme is proposed to solve the problems of privacy protection and utility in spatial range query.Specifically,considering the data storage and computing risks faced by the single cloud architecture,the outsourcing cloud of the scheme is designed as the dual clouds architecture based on the idea of divide and conquer,and the corresponding sensitive information is stored on two non-collusive clouds respectively.The structure has the ability to resist inside attacks while guaranteeing the confidentiality of outsourcing data.Besides,considering that there are eavesdroppers on the transmission channel,the scheme also improve the ability to resist eavesdropping attacks while guaranteeing users' query security via the structure of dual clouds.Through homomorphic encryption and searchable symmetric encryption,the scheme supports both range query and type query.On the basis of guaranteeing users' query security,the scheme effectively improves the pertinence of querying and reduces the computing burden of cloud side and client side.Theoretical analysis and experiments show the security and retrieval efficiency of the scheme.(3)In sociality mode,an efficient location-retrieval privacy preserving scheme is proposed to solve the problems of privacy protection and utility in the process of quering friends'locations.Specifically,considering the user-defined privacy preserving requirements,the scheme offers a new user-defined access control policy named grid cell access control policy.For the geographic locations or sensitive areas that users are reluctant to share with their friends,they can use grid cell access control policy to prevent privacy from leaking in social networks.To improve the retrieval efficiency in the process of quering friends' locations,the scheme designs a method named associated grids based on grid structure and users'query distances.Since the method can filter out the locations of friends that are not in the associated grids while protecting users' privacy,the scheme can reduce the burden of distance calculation and comparison on the location server,and improve the efficiency of location retrieval compared to the previous schemes.Cost and security analysis show that the scheme is a practical and effective privacy preserving scheme.The simulations also verify the effectiveness of the scheme.(4)In review mode,an enhanced review-guidance-effectiveness privacy preserving scheme is proposed to solve the problems of privacy protection and utility caused by users'reviews.Specifically,considering the identification or inference attack caused by the distribution and quantity of users' reviews,the scheme evaluates the possibility of being identified by formalizing the attacker model and defining the similarity probability among users.On this basis,the number of a user's public reviews can be determined combined with the user's privacy preserving threshold,which can reduce the risk of the identification or inference attack without affecting the user's enthusiasm for leaving reviews.To improve guidance-effectiveness of reviews and reduce the impact caused by fake reviews,the scheme adopts voting decision rule and beta reputation mechanism to establish a reputation system for users,and takes users'reputation scores as the basis for ranking reviews in review lists.Theoretical analysis and simulation results show that the scheme can effectively resist the identification or inference attack,and enhance guidance-effectiveness of reviews.