Encrypted Data Search Technology In Cloud Storage

In cloud storage, in order to protect data security and user privacy, a commonly used method is to encrypt the data. Different from the plaintext, encrypted data is difficult to operate and other users can’t access the data except the data owner. This seriously affected the availability of data, and the value of which has also been greatly decreased. In order to ensure data security and privacy of users and improve the data availability, we studied the LSH(Locality Sensitive Hashing) similarity search method in depth and proposed an E-LSH similarity search method on the basis of the characteristics of the massive amounts data in cloud storage. E-LSH adds pseudo-similar objects elimination on the second phase of LSH query process. Without affecting the accuracy of the query, we have improved the query efficiency and solved the data blocking caused by encryption.E-LSH method separates the index and data store, separates the feature item index and encrypted data, and transcodes the feature items. The user can’t access the data during the searching process, which can not only protect the privacy of the user’s queries, but also can to ensure the security of data. To meet the needs of massive data search in cloud storage, we implemented the E-LSH method with Map Reduce on Hadoop distributed environment. Our experiment results show that compared with Multi-Probe LSH method, E-LSH method reaches a speedup of 18.4%.In addition, we have also designed a cloud storage encrypted data search system. The system realized the searching and sharing of encrypted data on the basis of ensuring data security. The system supports user keywords query and data objects query. System builders can also be customized according to the different types of data, which can not only meet the user’s data searching and sharing needs, but also is convenient to the system builders and cloud storage service providers.