Research On Fuzzy Multi-keyword Retrieval Scheme Over Encrypted Data In Cloud Computing

With the maturity of cloud computing technology,cloud storage has emerged.A lot of people and enterprises have chosen to store their local data on the cloud server because cloud computing can efficiently reduce the cost of infrastructure.However,due to the cloud server would infer more information based on the data stored on it,users can not fully trust the cloud server.How to ensure data privacy on the cloud server is an important issue.Most users choose to encrypt the data before uploading,while the encrypted data cannot be searched as plaintext.Traditional retrieval scheme based on plaintext no longer applies.The existing ciphertext-based retrieval scheme which is less efficient and has more space overhead does not support fuzzy search,multi-keyword retrieval or sorting of search results.Therefore,it is of great significance to study safe and efficient ciphertext retrieval methods.Based on the above problems,this paper mainly proposed a fuzzy multiple-keyword retrieval scheme for encrypted data in cloud environment(FMRS).The cloud server can retrieve the encrypted data containing the specified keywords according to the user's need and support the fuzzy keyword retrieval,while it can realize the privacy protection and would not disclose any information related to the data or retrieval to itself and other attackers.The process of building secure index and retrieval is more efficient and the sorting results are more accurate by using counting Bloom Filter and Min Hash algorithm.By using the distributed computing model MapReduce to optimize the system architecture as a distributed parallel way and designing the index structure as a distributed organization,we can solve the problem that the retrieval efficiency would be reduced because of the increasing of the index size in the centralized environment.Experiments show that the proposed scheme has higher secure index construction and retrieval efficiency than the existing scheme,and has higher retrieval precision.It can also further improve the retrieval efficiency by using distributed parallel way.