Research On Searchable Encryption Schemes With Efficiency And Security For Cloud Storage

With the rapid development of cloud computing,more and more enterprises and individuals are willing to outsource their data for higher quality data services.In order to preserve the privacy of sensitive data,an usual approach is to encrypt data before outsourcing them into the cloud.However,when data is stored in the form of cipher-text,the relation between the original data is lost.Hence,it is hard for people to search over encrypted data.Searchable encryption schemes are widely applied to solve the above problem.In searchable encryption schemes,users can retrieve the data back through keyword-based search over encrypted data.However,most existing searchable encryption schemes cannot be applied directly to the cloud database,especially for relational cloud database model.When the database is deployed on the cloud storage in encrypted form,the value and the relation between data items are modified.Therefore,users are unable to execute SQL queries over encrypted database directly.In addition,with the popularization of mobile smart devices,people would like to perform data storage and management via their mobile devices.Although encryption can achieve privacy protection,it requires a complex cryptographic manipulation before decryption.Most existing searchable encryption schemes take up a lot of resources during decryption,resulting in quite high computational overhead.Since resources of mobile devices are limited,such as memory space,CPU processing capacity,the existing encryption schemes cannot be applied directly.In this thesis,we have mainly studied on searchable encryption schemes with privacy preserving for cloud storage.In view of the different security requirements of two application scenarios,two secure searchable encryption schemes are designed respectively.The main contributions are as follows:(1)We proposed an efficient searchable encryption scheme on the basis of the analysis of the security requirement of cloud database application scenario.In this scheme,users can perform SQL queries over encrypted cloud database through combining block encryption with multi-layer encryption technology.After specifying the keyword columns,user encrypts database with block encryption.Then user encrypts keyword columns into multiple encryption columns with multi-layer encryption.When the user executes SQL queries,the system adjusts the corresponding encryption column to the most secure encryption layer that enable to run the requested queries.Lastly,the performance analysis and experimental simulation results show that our scheme can effectively improve query efficiency,and also realize the confidentiality protection during query process.(2)In order to solve the problem of resource-constrained mobile devices,we proposed an efficient and secure searchable encryption scheme.The scheme we proposed supports fast decryption and fine-grained access control through combining with attribute-based encryption.In this scheme,we proposed a public-key encryption strategy based on bilinear mapping.We allowed the private cloud to participate in the partial decipherment so as to reduce the computational overhead on user's side during decryption.Meanwhile,we encrypt user's private key with cipher-policy attribute-based encryption(CP-ABE)to achieve fine-grained access control.Lastly,we gave the security and performance analysis of this scheme.The experimental simulation results show that the scheme can effectively reduce the computation overhead and resource consumption in the decryption process and greatly improve the query efficiency.(3)We designed and implemented the prototype system of multimodal visual-audio emotion recognition.The Qt and Visual C ++ were used to realize the prototype system.We implemented SQL query over encrypted data scheme and searchable encryption scheme based on bilinear mapping on this prototype system and the feasibility of the scheme was verified.