Article Preview
TopIntroduction
Cloud computing offers to users a lot of services, cloud storage is among these services. The owner can store in the cloud a lot of kinds of data like databases, text files, audio files, video files, photos, emails, and many other data. Owners choose to store data in the cloud to benefit from its various advantages (Chang & Wills, 2016). for the reason of security, data are encrypted before sending them to the cloud. Owners store data in the cloud and share it with users. However not all users are authorized to get these data, access control must be assured. Users can access authorized data or search to find authorized data that correspond to some keywords. Searching with multi-keywords is more suitable than a single keyword. To protect sensitive data, data retrieval must be done without any data leakage. Recent data storage and retrieve approaches reveal important information; like the number of matching documents; or don't assure access control or made only a single keyword search.
A lot of cloud storage works; like the work of Yang et al. (2019), Liang et al. (2019), Riad et al. (2019), and the authors of this paper; take personal healthcare data as an example and try to enhance its privacy in cloud storage. Sharing personal healthcare data between patients and doctors using cloud storage will facilitate medical examination and doesn’t require patient displacement; however sensitive personal healthcare data need secure share and storage because simple modification could cause catastrophic results. In addition to that, patients need to control access to their healthcare data; so only selected: doctors, nurses, researchers, and family members can access this information. Every proposed work treats the problem from its viewpoint; Yang et al. (2019) have thought about the access of the on-site first-aid personnel when the patient’s life is threatened and propose a secure system that makes self-adaptive access control for both normal and emergency situations, Liang et al. (2019) treat the problem of generating decryption key for authorized users and finally Riad et al. (2019) think about dynamic access where users can get dynamically changing permissions based on authentication and context attributes.
In the author's approach, two works were proposed: Ali Guechi and Maamri (2017) propose in the first work solutions to some limits found in the work of Henz et al. (2014). Authors focus on how to protect shared and stored data from unauthorized access. To achieve this goal, the authors propose to encrypt the data on a per-access basis and to improve security; the authors propose to destroy the encrypted data after a period to perform new encryption. In the second work, Ali Guechi and Maamri (2018) have inspired by the approach of Cui, H. et al. (2017) to develop new work. Authors focus on how to securely search encrypted shared data to retrieve encrypted data matching certain keywords while forcing access control and decreasing the revealed information. To improve their work, the authors propose to use multi-cloud to benefit from its various advantages.
In the author's approach; to assure access control; data items are encrypted by fields. To retrieve data that correspond to some keywords, authors use an inverted index; which is a kind of table. This is called searchable encryption, where the data owner encrypts his data and prepares an inverted index, and sends them to the cloud. A data user who wants to retrieve data that correspond to some keywords will send a trap (request) to the cloud, this one will use an inverted index before sending an answer to the data user. Ali Guechi and Maamri (2018) propose inverted indexes where access control is steel assured. They propose also to send data to multi-cloud which helps to reduce information revealed. The authors' approach differs from other approaches on different points: