Securing IoT Networks: A Deep Learning Strategy Against RPL Selective Forwarding Attacks
Ayoub Krari, Abdelmajid Hajami, Ayoub Toubi, Soukaina Mihi · International Journal of Engineering Trends and Technology · 2024
In the dynamic and rapidly evolving domain of the Internet of Things (IoT), the imperative to safeguard networks against sophisticated cyber threats, notably selective forwarding attacks, has become increasingly urgent. This research introduces a novel strategy leveraging the capabilities of a Multilayer Perceptron (MLP), a sophisticated form of feedforward artificial neural network celebrated for its pattern recognition efficacy, to significantly bolster IoT network security. Motivated by the escalating complexity and subtlety of cyber threats, this study aims to develop a robust model capable of discerning and neutralizing selective forwarding attacks with high accuracy. The methodology employed encompasses the emulation of IoT environments using the Cooja Simulator for comprehensive data acquisition, focusing on network attributes essential for effective MLP analysis. The preprocessing of this data, including normalization and missing value imputation, is critical to refining the dataset for optimal analysis by the MLP. The architecture and training of the MLP are detailed, emphasizing feature selection and hyperparameter optimization to mitigate the risk of overfitting while maximizing detection capabilities. The efficacy of the proposed model is validated through empirical evaluation, employing a suite of performance metrics such as accuracy, precision, recall, and the F1 score. These metrics confirm the model's effectiveness in distinguishing between benign network behavior and potential attack scenarios, underscoring its applicability to IoT network security. Additionally, the study considers the practical integration of the MLP model within real-world IoT infrastructures, addressing the unique challenges and operational demands of such networks. Given the continuous advancement of cyber threats targeting IoT networks, the urgency of this research is evident. The proposed MLP model not only demonstrates significant potential in detecting selective forwarding attacks but also serves as a scalable and adaptable framework for enhancing the security posture of IoT networks. This investigation contributes to the cybersecurity field by offering a potent solution for protecting IoT infrastructures against an ever-evolving threat landscape, thereby ensuring their resilience and integrity in the face of sophisticated cyber threats.