Secure and Efficient Multi-Attribute Range Queries based on Comparable Inner Product Encoding

Abstract

Encryption, a powerful tool for data security, has been widely applied to protect sensitive data stored on untrusted cloud servers. One important problem in such an environment is how to support advanced query predicates, such as range queries, over an encrypted data set in an efficient and secure way. Order-preserving encryption (OPE) produces ciphertexts that preserve the order of their plaintexts and performs range queries directly on ciphertexts. However, ideally secure OPE schemes are inefficient (interactive and stateful), because they either ask for extensive client-to-server interactions or require a large persistent client storage that relates to the size of the data set. In this paper, we propose a comparable inner product encoding (CIPE) scheme to support multi-attribute range queries over encrypted data. Our main idea is to encode data and query values as encrypted vectors so that order comparison is realized by calculating the vector's inner product. Compared with existing OPE schemes, our scheme has the following merits: 1) High $e$ ficiency. It allows a client to retrieve data of interest in one round without maintaining any local state. 2) Enhanced security. It achieves ideal security while effectively resisting inference attacks that existing OPE schemes are vulnerable to. Extensive experiments conducted on a real- world, large-scale data set verify the effectiveness of our scheme.