Identification of Cryptographic Algorithms Based on CNN
Hailiang Hu, Ke Yuan · 2025
In the digital age, information security is extremely important, and encryption technology is the key to ensuring data confidentiality. With the increasing variety of encryption algorithms, traditional methods for identifying encryption algorithms are facing great challenges. Therefore, this article proposes a deep learning based block cipher algorithm recognition scheme, using a convolutional neural network (CNN) model. By designing the feature extraction process and the powerful functionality of the model, this scheme can effectively identify five types of group encryption algorithms. The experimental results show that in the binary classification recognition task, the average recognition accuracy of this scheme reaches 87.5%. By optimizing the feature set through feature selection, the average recognition accuracy of binary classification tasks has been improved by 6.5%. This scheme improves recognition accuracy and provides strong support for dealing with complex and diverse encryption algorithms.