A Scheme of Relational Database Desensitization Based on Paillier and FPE
Miao Wu, Jun Sheng Huang · 2021 3rd International Conference on Machine Learning, Big Data and Business Intelligence (MLBDBI) · 2021
Data desensitization is a process that makes some kind of special transformations on sensitive data to keep the data secure. At present, relational database is widely used in various IT fields and the data stored in relational database can be queried efficiently by using indexes of database fields. With the rapid development of cloud computing, how to make the data in relational database secure and can be queried efficiently has become a problem to be solved. In order to resolve the problem of relational database desensitization, a scheme of desensitization is proposed. The improved FF1 algorithm which is called FF1-SM4 is proposed by changing AES to SM4. The general framework of scheme and the process of desensitization are designed. Paillier is utilized to encrypt numeral data and FF1-SM4 is utilized to encrypt character string. The scheme makes the data secure on the cloud, desensitized character string data can be queried using traditional SQL without decrypting and desensitized numeral data can be calculated in the form of ciphertext. The results show that the data security in relational database on the cloud can be maintained and can be used efficiently.