Framework for efficient search and statistics computation on encrypted cloud data

Abstract

Storing data in untrusted cloud, keeping it confidential and allowing search and other operations on the encrypted data without revealing any meaningful information is currently an area of major interest in cryptology. Various new encryption paradigms are proposed to address the problem. Hidden Vector Encryption (HVE) is one such approach that supports different kinds of queries (e.g. search, comparison etc.) as a conjunctive normal form. In this paper we present a framework for efficient searching and computing simple statistics on the encrypted database based on the functionality of HVE. We revisit the work of Tseng et al. from IWSEC'13 and identify certain limitations of their methodology. In this paper we propose several new encodings for different attribute classes that overcome the limitations of the previous work to a great extent. Our technique, not only increases the span of queries allowed but also improves the efficiency of most of the queries than in the work of Tseng et al. © 2014 Springer International Publishing.