Research On Ranked Fuzzy Keyword Search In Cloud Computing Environment

With the development of computer science technology, and the cloud computing technology becomes prevalent, more and more enterprises institutions and individuals prefer to store their local data such as customer information, personal health record and some other information into the cloud, since we can access data flexibly as well as reduce the local overhead costs. One of the important steps is that when one wants to access a file stored in the cloud server, he needs to search over all the files in the cloud to get the right one. Since the cloud server is not fully trusted, we encrypt the files before outsourcing them to the cloud. So how to retrieve files from the cloud server and get them back efficiently and securely becomes the focus of attention.Searchable encryption techniques in the cloud environment generally supports accurate keyword search, which neither allow spelling errors and format inconsistencies, nor capture the relevance among data files and return the files in a ranked order. Although privacy preserving fuzzy keyword search schemes and ranked keyword search schemes over encrypted cloud data have been developed in recent years, these schemes function unilaterally, and the search efficiency is not very satisfactory. Therefore, how to incorporate fuzzy keyword search with ranked keyword search and make it efficient and secure has become a hot topic in searchable encryption in the cloud computing environment.In this thesis, we first introduce the basic knowledge of cloud computing and the cryptography briefly, as well as the overall structure of the keyword search process. After that, we introduce some existing fuzzy keyword search algorithms and ranked keyword search algorithms over encrypted cloud data in detail.On this basis, we present a privacy preserving ranked fuzzy keyword search algorithm over the encrypted cloud data, which supports both fuzzy keyword search and ranked keyword search. On the other hand, this scheme has changed the hash function into locality-sensitive hashing to improve the accuracy of search. Finally, we simulate the scheme under Linux system multi-threading to improve the search efficiency, since the search process can execute concurrently in multi-threading. Besides, the scheme constructs index separately, so it is not necessary to construct index again if the files in the cloud server do not have add, delete, or modify operations, which can shorten the search time and improve the search efficiency.As can be seen from the simulation results, our scheme supports both fuzzy keyword search and ranked keyword search, improves search accuracy and efficiency, and ensures the search security compared with the existing keyword search technology.