Research on the Application of Neural Network in Electronic Information Encryption and Decryption
Year: 2024
Type: article
Abstract: This paper explores the application of neural networks in electronic information encryption and decryption, investigating how deep learning technology can be utilized to enhance the security and efficiency of encryption algorithms. The article first introduces the development history of neural networks and their promising applications in the field of information security. It then provides a detailed description of an encryption and decryption framework built by combining Convolutional Neural Networks (CNN) and Long Short-Term Memory (LSTM) networks. This framework employs CNN to extract features from plaintext data and uses LSTM networks to dynamically generate keys and initialization vectors (IV) for the Advanced Encryption Standard (AES) algorithm, thereby enhancing the randomness and anti-cracking capabilities of the encryption. Experimental results indicate that this method significantly improves upon traditional AES algorithms in terms of security, flexibility, and encryption strength. Additionally, it demonstrates advantages in automated feature extraction and reducing the workload associated with manually designing features. The research presented in this paper not only holds theoretical value but also has practical significance for safeguarding national information security and promoting the development of information technologyr.
Author: Yingqi Ding
Institution: Guangdong Polytechnic Normal University
Language: English
Subfield: Artificial Intelligence
Field: Computer Science
Domain: Physical Sciences
Open Access status: diamond
OpenAlex ID: W4405730032