Encryption technique based on fuzzy neural network hiding module and effective distortion method

Underrating the evergrowing effectiveness of cryptanalysis tools is likely to be a big mistake. Although available protection techniques are effective and may dissipate our worries about information safety, we should not ignore the formidable advancements in hacking tools. Such advancements are likely to soon challenge our protection techniques. Unfortunately, although this challenge is probable, there is not much advancement in the way the protection tools function. Encryption techniques still use roughly the same computation principles and the same level of dependency on the encryption key. This paper goes beyond these encryption techniques and proffers an encryption technique with a new computation model. This model uses a fuzzy neural network to generate highly complicated hiding codes from the encryption key. The computation model uses also substitution and distortion methods that depend on plaintext and chaotic noises to induce enormous confusion in the ciphertext. This combination along with other confusion boosting actions such as interval scattering and chaotic locking establishes a very effective encryption technique. Experiments on the proof-of-concept prototype showed that the output (ciphertext) of the proposed technique passed rigorous randomness tests.