Study On Secure Query Based On Location Privacy In Spatial Crowdsourcing

Spatial crowdsourcing(SC)is a crowdsourcing service model based on location information,whereby the task matching process is that worker will select tasks outsourced by the requester according to their work scope,and then go to complete the task in the required locations for paid,such as Meituan,Didi.The convenience service has actively promoted the development and popularization of spatial crowdsourcing,and also promoted related academic research.However,people are more concerned about their data privacy and security due to the personal information leakage cases in recent years,especially involved sensitive information such as home address and work location.Unfortunately,the existing spatial crowdsourcing privacy protection schemes have many shortcomings,such as one-side protecting,low task matching efficiency,or inability to dynamically update data.To achieve efficient task matching based on location privacy under spatial crowdsourcing,this paper will research the following two aspects:(1)Multi-user Secure Range Query scheme:To protect the location privacy security of Requester and Worker under spatial crowdsourcing,this paper first proposes a Multi-user Secure Range Query scheme(MSRQ)based on location privacy security of multiple users to achieve secure task matching.Among them,this scheme utilizes bilinear mapping method to protect location data privacy,and utilizes proxy re-encryption technology to realize the user revocation.Simultaneously,to realize the range query under the ciphertext,the scheme utilizes the segment tree to convert the task location information of the Requester and the range query information of the Worker,thereby effectively turning the numerical comparison process of the range query into a prefix matching process.(2)Secure query mechanism based on the dynamic index of segment tree:To improve the efficiency of task matching under spatial crowdsourcing,based on the proposed MSRQ,this paper designs a novel secure and dynamic tree-based index(SD-Tree)to shorten time cost of task matching.This method cleverly turns the index construction problem into an index merging problem.First,the segment tree is utilized to convert the task coordinates of the Requester into the index branch structure,and then encrypted and uploaded to the SC-server for merging into the index SD-Tree,thereby improving the performance times of query efficiency.In addition,the program also supports dynamic updates in logarithmic time.Finally,this paper uses the real-world location-sharing data set Gowalla to simulate experiments on the proposed method,and the results verify the usability and practicability of the design scheme.Besides,this paper has also compared with the related research approach SC-MSDE.The theoretical analysis and experimental results show that our scheme has obvious advantages in terms of task matching efficiency.