ABSTRACT
Due to advancement in cloud computing technology, the research on the outsourced database has been recently spotlighted. However, because the outsourced database may contain personal information of users, the private information can be expos...
ABSTRACT
Due to advancement in cloud computing technology, the research on the outsourced database has been recently spotlighted. However, because the outsourced database may contain personal information of users, the private information can be exposed by the malicious attackers, i.e., service provider. In order to resolve this problem, it is important to encrypt the database prior to outsourcing it to the service provider.
On the other hand, in terms of query processing, it is impossible to process a query over the encrypted database without decrypting it. However, it is time-consuming and inefficient way. Therefore, the order-preserving encryption scheme has been proposed to enhance the query processing efficiency. The order-preserving encryption hides the original database while allowing the order preservation of the encrypted values so that the service provider can perform a query processing without database decryption. In the order-preserving encryption scheme, the encrypted data is sorted by certain column values, so it is weak to both order matching attack and count attack.
Therefore, we, in this paper, propose a partial order-preserving encryption scheme and a query processing algorithm. First, our partial order-preserving encryption scheme divides database into data groups and encrypts the data in each group by using periodic function. Therefore, we can preserve the order of each data group, but not each data. So the proposed encryption scheme is robust to borh order matching attack and counting attack. Second, our query processing algorithm utilizes a signature for each data which stores the data original data information. By using signature information, we can efficiently perform the exact matching query and range query over the encrypted database.
Through our performance analysis, it is shown that the proposed algorithm outperforms the existing algorithm in terms of range query processing and exact matching query processing while providing the higher level of privacy protection.