Text File Privacy on the Cloud Based on Diagonal Fragmentation and Encryption

Abstract

Despite the growing reliance on cloud services and software, privacy is somewhat difficult. We store our data on remote servers in cloud environments that are untrusted. If we do not handle the stored data well, data privacy can be violated with no awareness on our part. Although it requires expensive computation, encrypting the data before sending it appears to be a solution to this problem. So far, all known solutions to protect textual files using encryption algorithms fell short of privacy expectations. Thus is because encrypting cannot stand by itself. The encrypted data on the cloud server becomes full file in the hand causing the privacy of this data to be intrusion-prone, thus allowing intruders to access the file data once they can decrypt it. This study aimed to develop an effective cloud confidentiality model based on combining fragmentation and encryption of text files to compensate for reported deficiency in encryption methods. The fragmentation method used the strategy of dividing text files into two triangles through the axis. Whereas the encryption method used the Blowfish algorithm. The research concluded that high confidentiality is achieved by building a multi-layer model: encryption, chunk, and fragmentation of every chunk to prevent intruders from reaching the data even if they were able to decrypt the file. Using the privacy accuracy equation (developed for the purpose in this research), the model achieved accuracy levels of 96% and 90% when using 100 and 200 words in each chunk on small, medium, and large files respectively.