Low retrieval consumption searchable encryption scheme based on random number pairing
Rongrui Chen, Hai Ying Cheng · 2023
With the development of cloud technology, cloud servers have become the choice for people's data storage. However, compared to local storage, cloud server storage also brings some security issues. To avoid cloud servers from stealing user data information, users need to encrypt the files and upload them. The user efficiency of this approach is not efficient, so the concept of searchable encrypted files is proposed. Most of the current searchable encryption schemes are based on bilinear pairing, and this approach makes keyword retrieval in encrypted data to be inefficient. Moreover, most of the schemes are not resistant to keyword-guessing attacks from inside the server. Therefore, we propose a method to improve retrieval efficiency by breaking homomorphic encryption decryptability combined with random number sharing. At the same time, it achieves CI (Ciphertext Indistinguishability) and TI (Trapdoor Indistinguishability) security. The ciphertext is combined with a trapdoor to generate a list of results instead of deterministic results as a way to avoid internal keyword guessing attacks. After conducting extensive experiments, it is shown that our proposed method has a relatively significant improvement over other schemes in terms of retrieval overhead and computational overhead of users.