Аннотация
Intrusion detection plays a critical role in protecting Internet of Things (IoT) networks from cyber attacks, such as Denial of Service (DoS), Man-in-the-Middle (MitM), and spoofing attacks. This paper, in this regard, proposes a robust Multi-Scale Convolutional Neural Network (CNN) as an enhanced method for intrusion detection. The proposed model has been trained on the RT-IoT2022 dataset and is evaluated for key evaluation parameters, such as accuracy, precision, recall, F1 score, area under the curve (AUC), and receiver operating characteristics (ROC). In addition, several deep learning (DL) models, i.e., recurrent neural networks (RNNs), gated recurrent units (GRUs), convolutional neural networks (CNNs), and long short-term memory (LSTMs), have also been implemented on the same dataset for comparison purposes. The proposed model outperforms all other models in comparison in terms of evaluation parameters, with an accuracy and F1 score of 99.32 % and 99.22 %, respectively. The proposed model has also been evaluated for reliability parameters, i.e., Cohen's Kappa coefficient and Matthews Correlation Coefficient (MCC). The optimal range results of the reliability parameters, i.e., a Cohen's Kappa coefficient of 0.99151 and an MCC of 0.99154, respectively. These results validate that the results of the proposed model are in accordance with the actual data.