Enhancing Side-Channel Attacks Prediction using Convolutional Neural Networks

A common type of cyberattack is the side channel attack (SCA), which affects many devices and equipment connected to a network. These attacks have different types, like power attacks such as DPA, electromagnetic attacks, storage attacks, and others. Researchers and information security experts are concerned about SCA targeting devices, as they can lead to the loss and theft of important information. Using deep learning (DL) techniques in SCA analysis can be an efficient tool for detecting the SCA. Many previous works have tried to carry out the SCA in order to mitigate the impact of these attacks, but they encountered difficulties in detecting the SCA, whether in selecting the suitable dataset or applying the most efficient machine learning or deep learning techniques for achieving high performance. Therefore, we developed in this paper a deep learning-based model to detect SCAs using a dataset related to power attacks (the DPAv4 dataset) and the Convolutional Neural Networks (CNN) algorithm to train and test the DPAv4 dataset. The findings of our experiments showed a significant improvement in the performance of DL-based techniques in the detection of SCA. The proposed CNN-based model achieved an accuracy of 0.814 in detecting SCA and a loss rate of 0.581.