A Study Of Hadoop-based Ciphertext Retrieval Key Technology

In recent years, with the rapid development of social networks, sensor technology, Internet technology, the amount of data generated is increasingly expanding everyday. It becomes a problem for storing data. It makes storage and retrieval of data possible through big data computing and cloud storage technology. Internet users can migrate their data from the local to the cloud, this can save infrastructure investment and data can be more efficiently managed. Because of the cloud itself openness and sharing, Security for data stored in the cloud face on a very big challenge. So the cloud data encryption is a common solution. For how to use the cloud data encrypted, a way is download the data and decrypt, this solution only suit for small data, but this way cannot be used for large data, so we need find a cloud searchable encryption technology to solve this problem. Data will not be exposed to anyone else and the user can retrieve keywords after the data is encrypted storage in the cloud. Currently, Searchable encryption technology available based on encryption type are divided into Searchable Symmetric Encryption and Searchable Asymmetric Encryption, for the structure of scheme, it is divided into a searchable encryption base on liner search and index-based searchable encryption scheme. These searchable encryption schemes have focused on safety or performance and can solve difference requirement of difference scenarios.This paper carried out extensive research on searchable symmetric encryption scheme, analyze some searchable encryption scheme proposed by researcher in recent years and complete the research as below.1. We proposed an improved security index-based approach in full-text search scenario. In our approach, we build our hypothesis. By using oracle model of this assumption, we prove the security of our approach. Meanwhile, with the assurance of retrieval security, we optimized the complexity of search to achieve a logarithmic level. so the approach is useful for big data environment.2.we propose a hadoop-based searchable encryption scheme, it make our searchable encryption scheme using in big data environment through Hadoop cluster computing. At the same time, we present hadoop-based process for building indexes and retrieval in detail. They can support our hadoop-based ciphertext retrieval scheme effectively3.To prove our approach,we implement some popular searchable encryption scheme and our improved scheme.we compared by the time for building indexes and retrieval.It prove that our scheme is similar as others in building indexes, but it prove that the retrieval time of our scheme is close to logarithmic time base on the number of trapdoor. Retrieval time is better than other scheme and is more suitable big data environment.