Abstract:

The widespread adoption of Fog computing has ushered in new possibilities for efficient data processing and reduced latency at the network edge. However, the integration of edge and fog devices into the computing ecosystem introduces security challenges that demand comprehensive solutions. In this research, we address security threats in the Fog computing environment and propose an enhanced multimodal biometric authentication system (MBAS) that leverages face, ear, and hand vein images. Feature extraction techniques using Independent Component Analysis (ICA) and Linear Discriminant Analysis (LDA) are applied to enhance the discriminative power of the biometric traits. To further improve the accuracy of authentication, we employ a feature fusion mechanism based on Grade Level, Multi-Objective Mode Optimization Genetic Algorithm (MOMGA) feature selection is used to select the most relevant and discriminative features for classification. The authentication process is performed using the K-Nearest Neighbors (KNN) classifier. The effectiveness of the proposed method is evaluated using a real-world dataset comprising face, ear, and hand vein images collected from a diverse set of individuals. Experimental results demonstrate that the proposed approach achieves superior authentication accuracy compared to conventional biometric systems. Additionally, the use of MOMGA feature selection enhances the model's generalization capability and improves the system's resistance to attacks.

Description:

Fog computing, multimodal biometric authentication, Independent Component Analysis, Linear Discriminant Analysis, Grade Level fusion, Multi-Objective Mode Optimization Genetic Algorithm