Research On Preference-oriented Multiple User Spatial Query Authentication

As the mobile devices become popular in recent years,there are more and more location-based service(LBS)application providers.These applications bring convenience to the daily life of users,as well as dealing with massive data and requests.From the view of social division and improving productivity,data collector(or data owner)and service provider belongs to different professional field.Therefore,in the big data era,it's an inevitable trend for data outsourcing.However,as they belong to different security entities,the trust problem between data owner,service provider and users becomes a key challenge in the big data era.It's a hot issue for users to secure that service providers provides good service based on the original data.That is,it's necessary for users to authenticate the query result provided by service provider.There already exist some query types in LBS field,such as range query,k nearest neighbor query,top-k query,Skyline query,etc.These query types satisfy users' needs in varying degrees.However,for the multi-criteria queries in multi-user scenario,which includes both spatial and non-spatial attributes,there are less focus in this area.In such scenario,there are multiple users not only locate in separate locations,but also have different anticipation for the query results,which complex the problem.This paper mainly focus on the query authentication problems in the spatial data outsourcing scenario,especially the authentication problem for multiple user spatial queries,and proposes effective solutions for the problem.Firstly,this paper defines a new query type,i.e.,preference-oriented multiple user spatial query(PMUSQ),and analyzes current query authentication methods in spatial database outsourcing scenario,including their design patterns,functionality,security,efficiency,etc.,and draw a conclusion on the existing problems in these research,to provide the corresponding solution.Secondly,this paper designs two query authentication schemes targeting on preference-oriented multiple user spatial query,based on MR tree and MRS tree respectively,to secure the correctness and completeness of the query result for users.The proposed schemes not only secure the service quality of the service providers,but only have high efficiency for query authentication of users.Lastly,to further reduce the cost of query result authentication,this paper designs a novel VO returning strategy based on probability,which users selectively ask service provider to return VO by calculating a reputation for service provider.The proposed strategy massively reduces the cost of result authentication while remaining high probability of query correctness.