Inner Product Encryption For Privacy Preserving Keyword Searching In Cloud

With the emergence of cloud computing, an increasing number of users store their privacy data in the cloud servers, so rapid retrieval in cloud becomes a challenge. Predicate encryption which is a hot research area in cryptography is suitable for data access control system and achieves the retrieval of encrypted data in cloud. As a new type of encryption system, it can make up for the shortcomings of the traditional encryption system, but it still has many problems, such as low efficiency, single design idea and so on. Most existing predicate encryption schemes only achieve adaptively payload-hiding secure, but can not guarantee the security of the search predicates. Therefore, how to design a predicate encryption to search a file rapidly is very important. In this thesis, based on the advantages of the hierarchical identity based encryption, the key-policy attribute-based encryption and the predicate encryption, the predicate encryption for privacy preserving and keyword searching is studied in cloud. The main results are given as follows:1. Under the double dual vector space, new efficient hierarchical inner product encryption schemes are introduced in the standard model. The first scheme can support a wide class of predicates and distribute the private key to the lower level users through the superior one to avoid the leakage of the circulation. Furthermore, its security is adaptively attribute-hiding under the Decisional Linear assumption(DLIN), which is more natural than those in the available. This scheme achieves constant-size public keys and its decryption process is only required 2 pairing operations and 1 multiplication operations. Based on the scheme one, the length of the ciphertext can be linearly increased with the dimension of a space. So, a modified scheme is proposed. It achieves constant-size ciphertexts and has the same security as the first scheme.2. An efficient inner encryption for keyword searching scheme in the standard model is proposed. The first scheme is a predicate-only encryption scheme which is a subroutine in large systems with searching function. And then it is applied to the school data security system and achieves fully secure under the Decisional Linear assumption. However, it does not provide specific data information for data retrievers. To overcome the shortcomings of the above scheme, the improved scheme is put forward. It achieves adaptive predicate-hiding and attribute-hiding against malicious user, adaptive attribute-hiding against malicious helper, and adaptive predicate-hiding and attribute-hiding against malicious private-key generator(PKG).