Order-Preserving and Efficient One-to-Many Search on Encrypted Data
Order-preserving encryption (OPE) is an useful tool in cloud computing as it allows untrustworthy server to execute range query or exact keyword search directly on the ciphertexts. It only requires sub-linear time in the data size while the queries are occurred. This advantage is very suitable in the cloud where the data volume is huge. However, the order-preserving encryption is deterministic and it leaks the plaintexts’ order and its distribution. In this paper, we propose an one-to-many OPE by taking into account the security and the efficiency. For a given plaintext, the encryption algorithm firstly determines the corresponding ciphertext gap by performing binary search on ciphertext space and plaintext space at the same time. An exact sample algorithm for negative hypergeometric distribution is used to fix the size of the gap. Lastly a value in the gap is randomly chosen as the mapping of the given plaintext. It is proven that our scheme is more secure than deterministic OPE with realizing efficient search. In particular, a practical and exact sampling algorithm for the negative hypergeometric distribution (NHGD) is first proposed.