A Method of Query over Encrypted Data in Database

Abstract

The encryption mechanism is an effective way to protect the sensitive data in database from various attacks. When the data is encrypted, the query performance degrades greatly. How to query encrypted data efficiently becomes a challenge. In this paper, a scheme to support query over encrypted data is proposed. Firstly, we extend two-phase framework to complete query, constructing different indexes for different data types. A new method to construct Bucket index for numeric data is proposed and bloom filter compression algorithm is used for the character string, which will be saved in database as a numeric data. We analyzed the false positive probability of bloom filter to get optimal parameters. Experiment results show the performance of our scheme has improved.