BopSkyline: Boosting privacy-preserving skyline query service in the cloud

With the widespread adoption of cloud computing, there has been great popularity of storing and querying databases in the cloud. However, such service outsourcing also entails critical data privacy concerns, as the cloud providers are generally not in the same trust domain as the data owners/users and could even suffer from data breaches. In this paper, different from most existing works that propose security designs for keyword search, we focus on secure realizations of advanced skyline query processing, which plays an important role in multi-criteria decision support applications. We propose BopSkyline, a new system framework for privacy-preserving skyline query service in cloud computing. BopSkyline is designed to not only ensure the confidentiality of outsourced databases, skyline queries, and query results, but also conceal data patterns (like the dominance relationships among database tuples) and search access patterns that may indirectly lead to data leakages. Notably, through a delicate synergy of key ideas on secure database shuffling and differentially private database padding, BopSkyline achieves a significant performance boost over the state-of-the-art. Extensive experiments demonstrate that compared with the state-of-the-art prior work, BopSkyline is up to 4.7× better in query latency and achieves up to 99.38% cost savings in communication.