Enhancing Cloud Security A Multi-Factor Authentication and Adaptive Cryptography Approach Using Machine Learning Techniques

The rapid adoption of cloud computing has led to significant concerns regarding the security and privacy of sensitive data stored in cloud environments. Authentication plays a pivotal role in protecting user data from unauthorized access. This paper presents a robust security framework integrating multi-factor authentication (MFA) and adaptive cryptography to safeguard cloud systems against evolving cyber threats. The proposed system utilizes a combination of passwords, conditional attributes, and fingerprint authentication to verify users, while employing AES, ECC + HMAC (SHA-512), HMAC-MD5 + PBKDF2, and Blowfish + HMAC SHA3-256 encryption algorithms for secure data protection. A key feature of the system is the file upload module, where users can securely upload files encrypted with AES. To access these files, users must authenticate either through fingerprint recognition or by entering a one-time password (OTP). Additionally, a prediction tab is integrated, allowing users to input specific values to predict potential attack types. The attack prediction system, powered by a Hybrid CNN-Transformer model, classifies threats and adjusts the system’s encryption strategy dynamically to enhance security. The proposed framework demonstrates resilience against various attack vectors, including brute force, spoofing, phishing, and credential stuffing, achieving a remarkable 96.8% accuracy in attack prediction. The system’s adaptive nature allows it to continuously evolve and strengthen security based on real-time threat analysis, thereby providing a highly effective solution for enhancing cloud security.

Keywords: Multi-factor Authentication (MFA), AES Encryption, Fingerprint Authentication, One-time Password (OTP), Machine Learning Techniques, File Encryption.