Machine Learning-Enhanced Attribute-Based Authentication for Secure IoT Access Control

Area of Science:

• Computer Science
• Cybersecurity
• Machine Learning

Background:

• Rapid growth of Internet of Things (IoT) devices necessitates advanced authentication.
• Traditional systems face challenges in balancing security, privacy, and efficiency in resource-constrained environments like Industry 4.0.
• Existing solutions often struggle with computational overhead and protecting sensitive information.

Purpose of the Study:

• To present the SmartIoT Hybrid Machine Learning (ML) Model for enhanced IoT authentication.
• To improve security and minimize computational overhead in authentication mechanisms.
• To provide a solution suitable for low-power IoT devices and Industry 4.0 applications.

Main Methods:

• Integration of Attribute-Based Authentication with a lightweight machine learning algorithm.
• Utilization of Random Forest classifiers for real-time anomaly detection based on user attributes, login patterns, and behavioral analysis.
• Incorporation of privacy-preserving Attribute-Based Credentials and Attribute-Based Signatures.

Main Results:

• Achieved 86% authentication accuracy, 88% precision, and 96% recall.
• Demonstrated an average response time of 112ms, suitable for low-power IoT devices.
• Significantly outperformed existing solutions in experimental evaluations.

Conclusions:

• The SmartIoT Hybrid ML Model offers enhanced security, privacy, and computational efficiency for IoT authentication.
• The model exhibits strong security resilience, efficiency, and adaptability for real-world applications.
• It provides a viable solution for securing critical sectors and Industry 4.0 environments.