SecEQP: A Secure and Efficient Scheme for SkNN Query Problem Over Encrypted Geodata on Cloud

Nowadays, location-based services are proliferating and being widely deployed. For example, a Yelp user can obtain a list of the recommended restaurants near his/her current location. For some small or medium location service providers, they may rely on commercial cloud services, e.g., Dropbox, to store the tremendous geospatial data and deal with a number of user queries. However, it is challenging to achieve a secure and efficient location-based query processing over encrypted geospatial data stored on the cloud. In this paper, we propose the Secure and Efficient Query Processing (SecEQP) scheme to address the secure k nearest neighbor (SkNN) query problem. SecEQP employs the projection function-based approach to code neighbor regions of a given location. Given the codes of two locations, the cloud server only needs to compare whether codes equal or not to check the proximity of the two locations. The codes are further embedded into an indistinguishable Bloom filter tree to build a secure and efficient index. The security of SecEQP is formally proved in the random oracle model. We further prototype SecEQP scheme and evaluate its performance on both real-world and synthetic datasets. Our evaluation results show that SecEQP is a highly efficient approach, e.g., top-10 NN query over 1 million datasets only needs less than 40 msec to get queried results.