A Privacy-Aware K-Nearest Neighbor Query Scheme for Location-Based Services

The advancement of spatial positioning technology and mobile Internet makes it possible for location-based services (LBSs), which provide users with personalized services by collecting and analyzing their location information. The
$k$
-nearest neighbor (kNN) query algorithm can be used in LBS by returning
$k$
locations closest to the target location. Due to the fact that LBS involves large amounts of private information and is usually outsourced to cloud servers, privacy issues have become increasingly prominent. In order to provide LBS services in outsourced cloud environments while protecting user location privacy, a privacy-aware kNN protocol under the dual cloud server model is presented. In the proposed scheme, location-related information is encrypted using a double-trapdoor public-key encryption algorithm and stored in a
$k$
-dimension (KD)-tree before being outsourced to cloud servers. A Euclidean distance protocol and a comparison protocol are provided as basic components to realize secure search, insertion, and deletion of the encrypted KD-tree by the cooperation of two cloud servers. Security analysis indicates that the proposed scheme ensures location and query privacy. Performance evaluation results demonstrate that the scheme is practical in terms of time and communication consumption.