Enhancing Cloud File Storage Security through Cryptographic Techniques

Abstract

This project addresses the critical issue of secure cloud storage by developing a system that integrates hybrid cryptography and behavioral biometrics-based two-factor authentication (2FA). As cloud storage is increasingly vulnerable to data breaches and unauthorized access, our research problem focuses on enhancing cloud security through user-adaptive, behavior-based authentication alongside robust encryption.The primary objectives are to (1) secure data through a hybrid cryptographic approach, combining AES for data encryption with RSA for key exchange, and (2) enhance user authentication with behavioral biometrics. Methodologies include local AES encryption of data before cloud upload, ensuring secure access across devices through a web interface. For authentication, machine learning models analyze user-specific interaction patterns, such as mouse movements and scrolling speeds, to develop a behavioral profile for each user. This profile allows the system to detect impersonation attempts effectively. Initial results show a reliable authentication accuracy with a low false rejection rate, balancing security and user experience. Continuous monitoring and regular model updates allow the system to adapt to evolving behaviors, ensuring long-term efficacy. This approach demonstrates a scalable, user-centric solution to cloud storage security, addressing both data protection and adaptive authentication.