IoT Network Security Framework to Classify RPL Attacks Using Deep Learning Algorithms

The Internet of Things (IoT) is an innovative concept that has become widely popular due to its numerous benefits. However, the Internet of Things (IoT) encounters numerous obstacles to ensuring security. RPL attacks pose a significant threat to the security of IoT networks, potentially compromising the integrity and functionality of interconnected devices and systems. Nevertheless, this growing dependence requires strong cybersecurity measures. We propose utilizing two deep learning models, namely GRU and LSTM-CNN. These models have been trained on the Contiki Cooja IoT simulator dataset. Furthermore, the SMOTE-ENN approach is utilized to balance the distribution of the attack classes in the dataset. The efficacy of the two proposed models is evaluated by taking into account their precision and F1-score. The findings show that the GRU and LSTM-CNN models exhibit 90% and 95% classification accuracies, respectively, in identifying RPL attacks.