Private boolean query processing on encrypted data

Abstract

Outsourcing the data to the clouds offers an opportunity to drastically reduce costs of storing and processing data. On the other hand, it deprives the data owners of direct control over their data and that introduces new privacy risks. Data encryption has been introduced to tackle the data confidentiality issue. However, data encryption also brings a new challenge of query processing over encrypted data. Recently, solutions for supporting query over encrypted data have been developed. However, they are either failing to support complex queries or insecure regarding certain security requirements (i.e. access patterns, query privacy). In this paper, we propose a novel privacy-preserving query processing framework to support boolean queries over encrypted data. Our framework utilizes Bloom filter and additive homomorphic encryption to systematically derive the query evaluation results in a privacy-preserving manner. We theoretically and empirically analyze the performance of the proposed protocols and demonstrate their practical values.