Microsaccade-Based Biometric Authentication: A Secure and Privacy-Preserving Cryptographic Framework

Microsaccades offer a unique, hard-to-spoof biometric trait that enables secure and continuous user authentication in sensitive digital environments. This paper introduces an innovative biometric authentication system utilizing microsaccades-small, involuntary eye movements as a distinctive identifier for secure user verification. The system is architected around three core entities: the central management authority (CMA), regional branch authorities (RBAs), and users. The CMA acts as a trusted entity, managing user registration, cryptographic key generation, and training a machine learning model on microsaccadederived features such as velocity, amplitude, and frequency. RBAs handle local data collection using eye-tracking devices, preprocess and encrypt microsaccade patterns, and securely transmit them to the CMA for authentication. The system operates in three phases: Initialization, where the CMA establishes cryptographic parameters (elliptic curve parameters and symmetric encryption functions); Registration, where users’ microsaccade features are captured, processed, and stored; and Authentication, where RBAs are the entities responsible for encrypting and authenticating real-time microsaccade data, which is then verified by the CMA's machine learning model and responded to with access control tokens.