Statistical analysis, ciphertext only attack, improvement of generic quasigroup string transformation and dynamic string transformation

• Main Authors: Hassan Disina, Abdulkadir, Jamel, Sapiee, Abubakar Pindar, Zahraddeen, Mat Deris, Mustafa
• Format: Article
• Published: American Scientific Publishers 2018
• Subjects: T Technology (General); QA273-280 Probabilities. Mathematical statistics
• Online Access: http://eprints.uthm.edu.my/5379/; https://doi.org/10.1166/asl.2018.11626

Algebraic functions are the primitives that strengthen the cryptographic algorithms to ensure confidentiality of data and information. There is need for continues development of new and improvement of existing primitives. Quasigroup String transformation is one of those primitives that have many applications in cryptographic algorithms, Hash functions, and Pseudo-Random Number Generators. It is obvious that randomness and unpredictability is the requirement of every Cryptographic primitive. Most of those string transformations have not been implemented properly neither do they have security analysis. Cryptanalysis of existing scheme is as important as building new ones. In this paper, generic Quasigroup sting transformation is analyzed and found vulnerable to Ciphertext-Only-Attack. An adversary can compute the ciphertext to get the plaintext without prior knowledge of the plaintext. Pseudorandom numbers produced with generic string transformation can be reversed back to the original input with little effort. Therefore the generic quasigroup string transformation is compared with recently introduced string transformation and it is expected to provide better randomness and resistant to ciphertext-only-Attack. The proposed string transformation is suitable to one-way functions such as Hash functions, and pseudorandom number generators to mitigate the vulnerability of quasigroup string transformation to Ciphertext-Only-Attack. While the dynamic string transformation increase the difficulty level of predicting the substitution table used. The algorithms will be compared in terms of randomness using NIST statistical test suit, correlation Assessment and frequency Distribution.