A Risk-based Adaptive Authentication Model for Mobile IoT
Asma Arab, Ghada Jaber, Abdelmadjid Bouabdallah · 2025
The Internet of Things (IoT) has become an integral part of modern life, with a wide array of devices and technologies seamlessly embedded in various domains. However, IoT security—particularly in authentication—remains a critical challenge. Due to their limited resources, IoT devices require lightweight and efficient authentication mechanisms tailored to their capabilities. Additionally, the heterogeneous nature of IoT, where devices have varying security requirements, calls for flexible and adaptive solutions. Mobile IoT networks face even more complex challenges, as the threat landscape can vary across different geographical areas, necessitating real-time adaptation of security measures. Despite numerous proposed authentication schemes, existing solutions fail to address jointly IoT constraints, mobility and heterogeneity. To address this issue, we propose in this paper the AAM-mIoT solution, a robust and risk-based adaptive authentication solution specifically designed for IoT architectures. AAM-mIoT dynamically selects the optimal authentication method based on both the constraints of IoT nodes and the environment. Our solution relies on cluster-based system architecture and relies on a Software-Defined Networking (SDN) controller to monitor the network, ensuring efficient and secure authentication across the IoT system. Extensive experimental evaluations demonstrate AAM-mIoT superiority to ensure security compliance satisfaction and energy efficiency.