Enhanced Unimodal Continuous Authentication Architecture on Smartphones for User Identification through Behavioral Biometrics

Enhanced Unimodal Continuous Authentication Architecture on Smartphones for User Identification through Behavioral Biometrics

연구 분야: Analysis

학회: 2023 2nd International Conference on Vision Towards Emerging Trends in Communication and Networking Technologies (ViTECoN)

Nowadays intruders are smart and intelligent to initiate various attacks on the smartphone's entry-point au- thentication methods to access the user's stored information on the smartphone. Accessing users' confidential information rather than personal information by attackers can create a monetary or non-monetary gain. Thus, continuous authentication was proposed on smartphones to recognize the user from the entry point to the end of the session. Even though, we noticed that earlier works on unimodal-based continuous authentication were limited to specific applications on smartphones. This can create a vulnerable gap to access the users' information on the smartphone. Thus, we propose an enhanced unimodal-based continuous authentication architecture for user identification on the smartphone. The proposed approach instantaneously chooses the unimodal behavioral biometrics based on the foreground application of the device and performs the user identification through a machine learning model. The experimental results pro- posed approach show better results than the existing literature.