A Privacy Protection Algorithm For K-nearest Neighbor Search In Cloud Outsourcing Environment

With the development of mobile Internet,location and geographic information has become a very important entrance,nearest neighbor query as a very important query method in location-based services,the problem of privacy protection has been studied by many researchers.But with integration of mobile Internet and cloud computing,more and more enterprises want to take advantages of cloud computing,they not only want to upload encrypted sensitive data for storage and management,and hope that the cloud platform can do some calculations.In this scenario,there have been some new problems,one of which is the k nearest neighbor query privacy protection in cloud outsourcing.K-nearest neighbor query in the cloud outsourcing environment has received more and more attention and research.Many schemes have been proposed,but some of the current solutions can't guarantee the security of the query user[1,5,7],and some search efficiency is too low to deal with large data sets[2,9],and some schemes can achieve a high security,but the calculation of overhead is too large[6],in short,there is no scheme can also protect the user's location privacy and the data owner's sensitive data,and cope with large data sets sub-linear search efficiency.2016 Wang et al[5] proposed the use of order-preserving encryption and R tree to achieve a large data sets K-nearest neighbor search scheme,but the scheme will link the encryption keys to user,so the privacy of user and data owner's will be no longer safe.By improving Wang's scheme,we achieve a scheme which can protect the privacy of the querying user and the data owner's,and achieve sub-linear search efficiency.In this article,the system structure of the scheme is introduced first,then the role analysis of the participants is carried out,at last,the parts of the scheme are introduced in detail and correctness and feasibility of the scheme are verified by experiments.