Convolutional neural network based key generation for security of data through encryption with advanced encryption standard

Ismail Negabi, Smail Ait El Asri, Samir El Adib, Naoufal Raissouni


Machine learning techniques, especially deep learning, are playing an increasingly important role in our lives. Deep learning uses different models to extract information from the data. They have already had a huge impact in areas such as health (i.e., cancer diagnosis), self-driving cars, speech recognition, and data encryption. Recently, deep learning models, including convolutional neural networks (CNN), have been proven to be more effective in the security field. Moreover, the National Institute of Standards and Technology (NIST) recommends the advanced encryption standard (AES) algorithm as the most often utilized encryption method in several security applications. In this paper, a crypt-intelligent system (CIS) capable of securing data is proposed. It is based on the combination of the performance of CNN with the AES, by substituting the key expansion unit of AES with a CNN architecture that performs the key generation. Our CIS is described using very high-speed integrated circuit (VHSIC) hardware description language (VHDL), simulated by ModelSim, synthesized, and implemented with Xilinx ISE 14.7. Finally, the Airtex-7 series XC7A100T device has achieved an encryption throughput of 965.88 Mbps. In addition, the CIS offers a high degree of flexibility and is supported by reconfigurability, based on the experimental results, if sufficient resources are available, the architecture can provide performance that can satisfy cryptographic applications.


advanced encryption standard; convolutional neural networks; cryptography; deep learning; embedded systems; field programmable gate arrays; hardware description language;

Full Text:



Creative Commons License
This work is licensed under a Creative Commons Attribution-ShareAlike 4.0 International License.

International Journal of Electrical and Computer Engineering (IJECE)
p-ISSN 2088-8708, e-ISSN 2722-2578