abstract
- Passwords are no longer a safe authentication, as they are susceptible to multiple sophisticated attacks. This vulnerability underscores the importance of adopting multi-factor authentication (MFA) as an additional security layer to supplement traditional password-based systems. Many MFA techniques have appeared over the years. The two most popular are one-time passwords (OTPs) and Biometrics. This paper proposes an innovative MFA system that combines audio, relationships, memorability, and device portability without requiring specialized hardware. The proposed system leverages knowledge-based authentication using audio, capitalizing on humans' superior capacity to recall voices compared to complex passwords. The user must upload several audios of people he knows and identify each person by name and a couple of other factors. To complete the authentication process, the users must correctly identify two audio files they uploaded. We evaluated the system's usability, memorability, and accuracy, and our results indicate satisfactory performance. Higher education students developed the project using the didactic 'Research-based learning' technique and improved their competencies and technical skills. © 2023 IEEE.