An architecture for secure searchable cloud storage

Cloud Computing is a relatively new and appealing concept; however, users may not fully trust Cloud Providers with their data and can be reluctant to store their files on Cloud Storage Services. This paper describes a solution that allows users to securely store data on a public cloud, while also allowing for searchability through the user´s encrypted data. Users are able to submit encrypted keyword queries and, through a symmetric searchable encryption scheme, the system finds all files with such keywords contained within. The system is designed in such a manner that trust from a public cloud provider is not required. The solution satisfies confidentiality of data; data integrity is maintained, file sharing is catered for and a user key-revocation scheme is in place. A further advantage of this approach is that if there is a security breach at the cloud provider, the user´s data will continue to be secure since all data is encrypted. Users also do not need to worry about Cloud Providers gaining access to their data illegally. The architecture of the system consists of two components, the Client side application and the Server application running on the compute cloud. The client side application performs all the security operations on the data. Along with saving and retrieving data from the Storage Service, the Server application performs the processing involved in handling the encrypted queries. The performance overheads of such a system are potentially significant in terms of additional processing time and the size of the additional meta-data needed. Preliminary results show that the storage overheads remain fairly constant as input file sizes increase - as file sizes were increased from 3Kb to 147Mb, the security overhead remained between 1038b and 1053b. This overhead is basically insignificant when storing large files. Overall the benefits of a searchable encrypted cloud service are significant and the approach is viable for using public clouds while still re- - taining control of the data.