Residual Convolutional Network for Detecting Attacks on Intrusion Detection Systems in Smart Grid

Smart grid provides several benefits, such as reliability and affordability. Despite its benefits, this network has several shortcomings, including a lack of security. DoS attacks are considered one of the main damaging cyber-attacks on these networks. For this purpose, several techniques have been proposed to detect such attacks. However, the majority of these techniques deal with high rates of false alarm and misdetection. In the last few years, deep learning techniques, particularly convolutional neural networks, have received a great interest for detecting cyber-attacks. This study proposes a convolutional neural network-based technique, a residual neural network with 50 layers. In this technique, the tabular data are changed into images to improve the performance of the model. The results show that the proposed model outperforms other ML models and achieves an accuracy of 98.1%, a probability of detection of 99.12%, a probability of misdetection of 0.88%, and a probability of false alarm of 1.03%.