الفهرس 
IntroductionOnly 14 pages are availabe for public view
 |  | from | 125 |  |  |
| | | from | 125 | | |
Abstract
Computing Cloud storage services allow users to reach their files anywhere and anytime in an efficient private manner. Recently, many users prefer using the cloud to relieve the local storage burden. However, storing files on remote servers may cause a source of concern where these files may be hacked by unauthenticated users or by curious administrators. An intuitive solution for the user is to submit files into the cloud storage after encrypting them. During searching on certain information across large encrypted files, the user must download the full file and decrypt it to get this information. The previous solution leads to computation overhead. Many types of research tried to debug this issue in many ways. Some of them were concentrated on changing the encryption methods while others were concentrated on indexing/querying the encrypted data. Searchable encryption is a cryptographic primitive that allows users to keep sensitive data safe on the server-side while still being able to query the data, even if the data is encrypted without decrypting. To keep the searchability on the server-side, the so-called trapdoor is generated by the client and sent to the server. With the trapdoor help, the server is able to perform the search on behalf of the client on the still encrypted data without leaking information stored in the original data. This thesis introduces a multi-server searchable symmetric encryption (SSE) scheme called the Multi-Server Searchable Data Crypt “MS-SDC” that works on achieving a trade-off between efficiency and security. The proposed scheme has the merits of dividing the uploaded file in an encrypted form into blocks and distributing them across several storage providers where each server only holds a subset of file/block which is more acceptable than uploading the whole file directly to a single server. This method ensures more security for the information stored in the file. Besides this, it is flexible so that it enables the user to upload either an encrypted file with its keywords or a plain file. If the latter is uploaded, it extracts keywords for the uploaded file to be used afterward by the search engine to give the user the ability to browse across his own files. The file keywords are stored in a hashed form to hide any data access pattern on the server-side, as over time, the server can know some of the file content through dealing directly with file/query keywords. In both cases, the user has the ability to query/search for his encrypted file on the server-side without decrypting. Furthermore, there are various features that proposed different from these presented by previous works as the scheme is developed as a multithreaded-application to speed up the uploading time, and a unique master key is generated randomly for each uploaded file unlike the previous techniques where a single master key is created randomly for all the uploaded documents leading to easily hacking the system with master key leakage. Finally, the MS-SDC system is distinctive in smooth usage and robust where it can run with any browser and can be applied to any type of file. The experimental results demonstrated the effectiveness of our proposed system in comparison to previous works in terms of file uploading and searching time in addition to providing many new features such as applying many layers of security and keeping high-speed performance in an efficient manner. To support the usefulness of our proposal, the simulation results have proved the functionality and the ability to reduce the file upload time up to about 64% of the current research uploaded time via using multithreading implementation of the block distribution function.