SEDB: Building secure database services for sensitive data

Abstract

Database outsourcing reduces the cost of data management; however, the confidentiality of the outsourced data is a main challenge. Existing solutions [9,13,16,17] either adopt multiple encryption schemes for data confidentiality that only support limited operations, or focus on providing efficient retrieval with problematic update support. In this paper, we propose a secure database outsourcing scheme (SEDB) based on Shamir’s threshold secret sharing for practical confidentiality against honest-but-curious database servers. SEDB supports a set of commonly used operations, such as addition, subtraction, and comparison, and is among the first to support multiplication, division, and modulus. We implement a prototype of SEDB, and the experiment results demonstrate a reasonable processing overhead.