An Efficient Oblivious Random Data Access Scheme in Cloud Computing

With the development of cloud computing and cloud storage techniques, much attention has been focused on the privacy protection of outsourced data. Existing searchable encryption solutions can ensure the confidentiality and availability of data stored on the cloud. However, searchable encryption is vulnerable to statistical inference attacks, which exploit the disclosure of access patterns on encrypted indexes and encrypted file sets, which has become a potential way to reveal user privacy. Oblivious random access memory (ORAM) is an important means of concealing access patterns, yet its direct use in searchable encryptions is expensive. This paper presents a scheme for efficient and oblivious access to encrypted databases through encrypted indexes. This scheme is a hybrid ORAM scheme, which utilizes semi-homomorphic encryption to perform calculations in the ciphertext domain, overcoming the limitations of the huge overhead associated with Path-ORAM. For excessive amounts of data, semi-homomorphic encryption can significantly reduce communication and storage overhead. Our scheme can achieve high-security encrypted search and update operations at the same time. Moreover, the execution speed of ODS-Tree is 2-8x faster than that of ORAM-based schemes. In addition, the proposed scheme reduces the data block transmission and storage costs compared to existing frameworks.