Research On Privacy-Preserving In Location-based Services

Mobile Internet has revolutionized the concept of network services,and it reduces the cost of obtaining and sharing information for the individual.Dynamic location information is fully used to promote the rise of the location-based services.Since location information can be analyzed both quantitatively and qualitatively as an important factor,privacy issue is inevitable when mobile user exposes his private location to a location service provider who is untrusted.Massive studies present privacy-preserving methods to balance QoS and privacy based on “zero-sum game”.Location k-anonymity,location obfuscation and private information retrieval are the dominant technologies in LBS privacy-preserving.However,many contentious questions about the location privacy remain to be answered owing to the complex scenarios in LBS.For example,how to balance the conflict between user experience and privacy? It is require specific techniques to achieve location privacy-preserving in special scenarios.Different user privacy requirements and different privacy content lead anonymous region to be short of stabilization and consistency.Geographical space constraints are restrict public freedom,however,could become the pivot of privacy invasion.To solve these problems,considering the defect of traditional location k-anonymous and location obfuscation,the dissertation has carried on the improvement,optimization and further research.The main contents and contributions of this dissertation are indicated below.First,users periodically update locations to LBS server,which can lead to high network overhead,the paper proposed a location privacy-preserving model based on the centralized architecture and a location privacy-preserving scheme based on pseudo-random permutation.This scheme build on k-anonymous and private information retrieval technology,using the random encryption technology to ensure the security of the private data,to achieve the perfect anonymous,location blind query and on-demand query targets.It is proved to have untrace-ability,unlink-ability and resisted continuous queries attacks or inference attacks.Second,aiming at the problem of personalized service in geo-fencing,the paper proposed a LBS privacy-preserving method oriented to triggered query.This method follows the concept of location obfuscation and location k-anonymity,draws lessons from the secure tripartite computation to design triggered query protocols,solved the problems of location anonymoushidden,secret identity authentication and digital check-in,achieved privacy-preserving of active push service in LBS,satisfied the anonymity,untrace-ability and unlink-ability.Third,a central server is the system bottleneck and the anonymous group lack of mutual trust,the paper proposed a multi-anchor location privacy-preserving method without central server and user cooperation.The method uses multiple incremental search technology to expand the hidden area of query location,increasing the difficulty of the attacker inferring user's location.In order to meet the multiple incremental searches,this paper presents a novel method for nearest neighbor queries,which handle respectively the x-coordinate and the y coordinate,and eliminates quickly a number of POIs to limit nearest neighbor query result set within a smaller range.Fourth,aiming at the problems of the actual barriers in load network and the lack of mutual trust between mobile and LBS server,the paper proposed two kinds of regional query privacy-preserving methods for LBS.The former designed a novel framework to support location-dependent queries in road network.Based on the theoretical work of k-anonymity and Voronoi diagrams,it constructed a restricted set(k,r,s)to meet the requirements for location k-anonymity and location l-diversity,and implemented an efficient algorithm for continuous region query processing.The latter draws on the theory of secure two-party computation,using the oblivious transfer protocol,and proposed a location privacy-preserving method in the non-collaboration distributed structure.This method divided the common query into the query Q and the query Q' to achieve a global anonymity and query diversity,which not only guarantee the user's privacy but also protect the security of the LBS service information.