A Federated Learning Mechanism for Mitigating Selective Forwarding Attacks in RPL-based Internet of Things

Security is a significant concern in the Internet of Things (IoT), particularly routing attacks causing severe damage due to information loss. The paper introduces a Federated Learning (FL) (using MultiLayer Perceptron (MLP)) detection mechanism to address selective forwarding attacks in Routing Protocol for Low-Power and Lossy Networks (RPL)-based IoT networks. This approach utilizes a lightweight model that incorporates the IoT Routing Attack Dataset (IRAD) to improve the ability to detect routing attacks. These attacks can compromise the overall stability and security of the IoT networks. The comparative findings demonstrate that the proposed MLP classifier exhibits high accuracy during training and achieves the most efficient runtime performance. The suggested system demonstrated exceptional performance, achieving a prediction accuracy of 98.33%, precision of 98.67%, recall rate of 97.33%, and an F1 score of 98% on average. It outperformed the current leading research in this field.