Research And Implementation Of Database Ciphertext Query In Cloud Environment

With the rapid socio-economic development, the data generated in everyday business situations witnessed a gradual increase in recent years. The cost of the local storage management becomes expensive correspondingly. Cloud storage services can provide users with unlimited storage resources, and it can support the expansion-on-demand for storage capacity dynamically. Low cost, easy manage, flexible manner are characteristics of the cloud database. More and more companies choose to use cloud database to solve data storage issue. However, as the database service provider cannot be trusted, it is very hard to ensure the security of users’privacy data in the cloud. Security issue always restricts the development of the cloud storage technology. Client encryption is the most effective way to solve this problem, but it will results in the loss of some inherent data characteristics, such as orderliness and comparability. Client encryption will also lead to performance degradation in data retrieval, and it will seriously affect the efficiency of cloud database. Aiming at the above problem, the ciphertext retrieval for numeric and character data in database is researched in this paper.As data encryption leads to the loss of orderliness and comparability, range query become a difficult problem in the ciphertext retrieval of numeric data. This paper presents an improved strategy based on numeric order replacement. This strategy improves its equivalent retrieval based on a new encryption scheme. In the sequence generation process, HMAC function is used to improve security. For the range query, this paper proposes a bi-direction search algorithm. It can reduce unnecessary data transmission and decryption effectively, cause further decrease in the interactions between the local server and cloud database, and improve the query efficiency finally.For character data, encryption makes the position relationship between the characters difficult to distinguish. The efficiency of fuzzy query over encrypted character data needs to be improved. This paper presents a security enhanced ciphertext retrieval model for character data. In the index generation process, the capacity to resist statistical analysis attacks is enhanced by the hidden character frequency. In the generation process of position index, this paper uses a hash function set to avoid the situation in which same plaintext will generate same index, so the similarity analysis attack can be resisted. In the query phase, this paper proposes a two-stage ciphertext retrieval method based on two kinds of indexes. In this method, both the alphabet index and the position index are used in the filtration to improve the filtration efficiency of the initial phase.Based on DAS service model, this paper uses the cloud database service provided by Baidu to complete the implementation of ciphertext retrieval technology for character data in the cloud storage environment. Combined with the source code, this paper gives a detail introduction to some core modules in the system, such as the user key generation module, data encryption and decryption module, index generation module, and the ciphertext retrieval module etc. The system is tested from memory rate and retrieval performance in this paper. Except the security enhancement, the feasibility and efficiency of the ciphertext retrieval model for character data is proved.