An RF-DNN-Based Approach for Detecting Cyber Attacks in IoT Network

The rapid expansion of the Internet of Things (IoT) has encouraged hassle-free connectivity and automation across domains, but in parallel, it has also brought forth serious security issues. As they do not possess computing capability as well as consistent security mechanisms, IoT devices are being targeted with advanced cyberattacks, such as botnet attacks like Mirai and Gafgyt. As countermeasures, this work discusses the implementation of both machine learning (ML) and deep learning (DL) algorithms in efficient and effective attack detection based on IoT. Some conventional ML models such as logistic regression (LR), random forest (RF), support vector machine (SVM), decision tree (DT), K-nearest neighbors (KNN), and Naïve Bayes (NB) are contrasted with DL models such as artificial neural network (ANN), deep neural network (DNN), convolutional neural network (CNN), and long short-term memory (LSTM). To further augment detection, the RF-DNN hybrid model is introduced. Models are trained on and tested from the BotNeTIoT-L01 dataset, and model performance comparison is conducted using standard evaluation metrics such as accuracy, precision, recall, F1-score, and specificity. The proposed hybrid model performs superbly with 98.6% accuracy, 98.4% precision, 98.3% recall, 98.35% F1-score, and 98.8% specificity, outperforming all the baseline ML and DL models. These results confirm that the hybrid model combines the strengths of both paradigms extremely well, offering an extensible as well as scalable solution for IoT security monitoring in real time.