Optimized Biometric-Based Anonymous Authentication for Secure Mobile Edge Computing
E. Dinesh, R Akash, R Bhagavathi, Harrish Sundar A · 2025
We propose a novel biometric-based anonymous authentication system for facilitating operation in mobile edge computing networks. Hence, the proposed approach incorporates FaceNet to perform face embedding extraction for better reliability and Federated Learning (FL) for privacy-preserving model training. Therefore, dynamic learning rate optimization makes the system adjust effectively to the fluctuations in data distribution enhancing the models. An Android-based application acts as a front-end through which users can safely take and send templates to the back end for real-time processing. The performances obtained from experiments show a very high identification accuracy as well as effectively safeguarding users’ privacy; therefore, making the realisation of the system well suited for applications that requires high levels of privacy in edge computing.