A Study On Homomorphic Eccryption Retrieval Scheme Based On Cloud Storage

With the rapid development of the Internet and Computer Application, both of the individuals’ and enterprises’ data grows increasingly, which big data becomes a development tendency in informatization. The rise of cloud storage service has brought a big data storage solution with high expansibility, undemanding management and low cost for individuals and enterprises. But as a third-party platform, due to its openness and sharing, cloud storage server may have the threat of privacy leakage if user’s data is uploaded without preprocessing. Cloud service provider or hacker can illegally acquire or access a user’s protected data, which has been a big challenge for popularizing cloud storage. An effective solution is encrypting data before storing it. Even if attackers get the ciphertext, they cannot retrieve useful information without the decryption key. But encrypted data lose usability, which is a problem to process data. Because we cannot directly conduct the update operation to the content such as adding, modifying, deleting the data in a file, the resources of cloud computing can rarely be put into efficient use. Traditional encryption limits the development of data application and hinders data management in cloud storage. It is significant for promoting cloud storage security application to study how to retrieval information from ciphertext effectively when users want to search what they need in massive data storage.Homomorphic encryption is an encryption that can perform direct compution on ciphertext, yet get meaningful result. The technique of homomorphic encryption can ensure data security and cipher usability, and there will be a great potential for the development in cloud storage environment. Based on researching a large number of related homomorphic theories and understanding of the technical features, this thesis analysizes the homomorphism in traditional public key encryption algorithm, such as RSA, ElGaml, Paillier, and elaborates on homomorphic encryption algorithms of the DGHV and CAFED.Through the study of the two homomorphic encryption schemes abroad, we analysize the safety defect during the procedure of ciphertext retrieval in which the original homomorphic encryption algorithm applies, and design an improved algorithm aiming to adapt to implementing the safety of ciphertext retrieval in cloud storage.Based on the improved homomorphic encryption retrieval scheme, we design a retrieval service model for cloud storage environment. Firstly, the model constructs the framework of the system. Then we start the detail design from the client, the management server and the storage server, which mainly consists of authentication mechanism design, encryption/decryption mechanism design, load balance mechanism design, virtual disk management design, homomorphic retrieval design. The techniques of authentication and virtual disk ensure access control and data segmentation; data encryption/decryption keep user ’s privacy; homomorphic retrieval achieves retrieval function as well as data confidentiality; load balance improves system’s flexibility and feasibility. Finally we’ve implemented the creation and loading of virtual disk, ciphertext storage and ciphertext retrieval, core functions of the system associating all modules. The safety of the service model has been proven through analysis of implementation process.