Hybrid Neural Cryptography: A Novel Approach for Enhanced Security and Efficiency in Resource- Constrained Systems

Abstract

The advent of artificial intelligence and deep learning has revolutionized multiple fields, and cryptography is no exception. Traditional cryptographic algorithms, such as symmetric and asymmetric encryption, offer robust security but come with significant computational overhead, making them unsuitable for low-resource environments such as Internet of Things (IoT) devices and embedded systems. On the other hand, neural networks have demonstrated the ability to perform complex transformations that are difficult to reverse, providing potential solutions for cryptographic systems. In this paper, we propose a Hybrid Neural Cryptography (HNC) model that integrates classical cryptographic algorithms with artificial neural networks to improve both the security and computational efficiency of encrypted communication systems. We present the architecture of the HNC model, its cryptographic components, and evaluate its performance against traditional methods, demonstrating its effectiveness in enhancing security while maintaining low computational cost.

DOI - https://doi.org/10.65525/SVUP.9788199565401.2026.177-183