Threats and Mitigation Strategies for Electroencephalography-Based Person Authentication.

Abstract

This work is aimed at investigating the potential risks linked to electroencephalography (EEG)-based person authentication and providing solutions to mitigate these issues. Authenticating a person by EEG involves verifying the legitimacy of the signals used for user identification. EEG signals serve as a biometric modality for authentication and verification. Additional biometric modalities, such as fingerprints or irises, are vulnerable to both fabrication and degradation over time, and illicit use of dead people's biometrics has sometimes been seen. EEG's intrinsic properties prohibit signal imitation or postmortem acquisition, making it more reliable than other biometric modalities. This research is aimed at investigating the most recent advancements in the domain of EEG signals, clarifying the current knowledge that impacts EEG-based authentication, and evaluating the emerging challenges. Many research publications have been collected to achieve this objective. By considering historical and recent efforts and achievements, this research also presents feasible resolutions to the emerging inquiries prompted by the ongoing advancements in EEG-based technology. The potential future application of EEG-based authentication has also been the subject of this scholarly discourse. A comprehensive collection of articles over the previous decade has been compiled to answer contemporary EEG signal research questions to get valuable insights. According to research findings, in February 2022, a significant milestone was achieved when the EEG signals of a deceased person were successfully captured for the first time in recorded history. However, this groundbreaking discovery may threaten EEG-based authentication. In addition, it is found that EEG-based authentication literature did not completely implement "liveness detection." An updated approach for identifying liveness addresses novel challenges, that is, falsified EEG signals and a dead person's EEG signals for EEG-based authentication that have not been discussed in the literature. The suggested solutions put forward in this study have the potential to stimulate additional research in this area.