Robust encryption technique based on a block-lag-induced reactive substitution, fuzzy neural network, and memory-loss stochastic resonance

Robust encryption technique based on a block-lag-induced reactive substitution, fuzzy neural network, and memory-loss stochastic resonance

연구 분야: Cryptography

학회: Neural Computing and Applications

Encryption techniques are becoming more robust through innovative advancements, but cryptanalysis techniques are also progressing. Although encryption techniques have one step ahead, this balance may shift as literature reports more breaches in encryption techniques. Current encryption methods are founded on innovative concepts like chaos, randomness, DNA, and neural networks; however, these alone may not provide enough protection against modern cryptography analysis. To effectively thwart cryptanalysis tools, encryption techniques should include methods with a global state to increase nonlinearity and confusion of ciphertext. This paper presents a novel encryption technique that integrates principles from chaos, neural networks, cognitive theory, and deep bit manipulation. The proposed technique can detect variations in the key and plaintext, and impose significant changes to the ciphertext. Additionally, it employs a reactive substitution space that monitors changes in the plaintext and adjusts the processing model accordingly, resulting in a highly-confused output. Experimental results show that the proposed technique meets rigorous security standards while maintaining efficient execution times, outperforming state-of-the-art techniques.