SSS-EC: Cryptographic based Single-Factor Authentication for Fingerprint Data with Machine Learning Technique

The development of cloud computing technology and big data comprises more users to store their data in the cloud server. The increases in the data volume and storage are subjected to the increased risk of data access with unauthorized users. Traditionally, to improve data authorization the cloud data is encrypted before uploading to the server. To improve cloud authentication Single Factor Authentication (SFA) techniques are evolved. However, conventional SFA is not efficient for sensitive information that is able to be accessed by third parties. To overcome this limitation, this research proposes a Single-factor Samoa Substring Escrow Cryptography scheme (SSS-EC). The proposed SSS-EC model uses fingerprint biometric data for authentication in cloud data. Initially, Samoa Substring is implemented with the validation of the client single-factor i.e fingerprint data. The validated information is stored in the cloud escrow. The validated data is encrypted using homomorphic encryption. The encrypted data is accessed with the attribute structure those need to query and decrypt the data in the Samoa Substring. Upon the verification of the attribute i.e., fingerprint, cipher text based on Samoa Sub-String is shared between the owner and user without any keyword. The verification with the cipher text is performed with Elliptical Curve Cryptography (ECC). The implementation of the SSS-EC scheme improves authentication in the cloud. Finally, the Machine Learning (ML) method is implemented for the classification of the different attacks in the cloud server using CICIDS dataset. The simulation analysis of the proposed SSS-EC model with the existing authentication techniques such as Ring Learning with Errors (R-LWE) and Identity Concealed Authentication Scheme (ICAS) based on two factors is performed. The proposed SSS-EC exhibits higher authentication accuracy and reduced computational cost for the different users and cloud servers. The experimental results confirmed that the pr..