Analysis And Constructions Of Public Key Encryption With Ekyword Search

With the profound revolution taking place in the field of information, Cloud Computing provides the demanding and scalable information service via combing a multitude of computing sources, storage sources and software sources. As one of the essential services in Cloud Computing, cloud storage allows the enterprises and individuals to outsource their to the cloud server and the cloud service providers manage and maintain the data as the principle of pay-as-you-use. However, even though cloud storage contributes to enhancing efficiency and inducing cost for corporations, it also triggers some significant challenges on data storage and privacy preserving simultaneously. Even though encryption technique has become an important approach to prevent the sensitive cloud data, it also leads to a big obstacle for searching and retrieving in encrypted data.Therefore, how to construct searchable encryption schemes on cloud data to address the conflict between the data encryption and search has become a challenging issue, which is the motivation of this thesis.In order to address above problem, this thesis will employ the technique of public key encryption with keyword search(PEKS) to meet the requirement of searching on ciphertexts efficiently, including discussing the security of some existing public key encryption with keyword search schemes, and constructing two efficient and practical encryption schemes with keyword search schemes combined with some new cryptographic techniques. Specifically:1. The security of two searchable public key encryption schemes with a designed tester due to Hu et al. is analyzed. It is demonstrated that these schemes are vulnerable to the off-line guessing attack, that is, a malicious server can distinguish the keyword that the user is searching via guessing the keyword in trapdoor. Then, we discuss the impossibility to construct secure PEKS with a designed tester schemes against off-line guessing attack, if the candidate keywords are from some polynomial size of dictionary.2. The notion of public key encryption with revocable keyword search(PERKS) is proposed to restrict the search power in cloud storage by revoking the search capability of the server. According to the anonymous multi-receiver identity-based encryption scheme due to Fan et al. and the approach of transform the anonymous identity-based encryption to PEKS presented by Abdalla et al., a novel and efficient PERKS scheme is shown combing the Lagrange Interpolation Polynomial and the method of dividing the whole life of the system into distinct time period. As far as we know, our scheme is the first one to provide a solution to revoke the capacity of the server’s searching in searchable encryption scheme.3. We propose a PEKS scheme supporting dynamic group from Zhao et al.’s dynamic asymmetric group agreement, Canetti et al.’s proxy re-encryption scheme and Boneh et al.’s classical PEKS construction. In order to reduce to burden of computation for users, a technique called server-aided computing is exploited to outsource the operation of bilinear pairing to the server. This new scheme meets the requirement of data sharing among users, users’ joint and revocation, identity privacy-preserving, keyword search and multi-contributors, and is adapted to cloud environment, particularly mobile cloud storage.