Proskurin, D. and Okhrimenko, T. and Gnatyuk, S. and Zhaksigulova, D. and Korshun, Natalia (2024) Hybrid RNN-CNN-based model for PRNG identification Classic, Quantum, and Post-Quantum Cryptography 2024, 3829. pp. 47-53. ISSN 1613-0073
![]() |
Text
D_Proskurin_T_Okhrimenko_N_Korshun_CQPC_2024_FITM.pdf Download (1MB) |
Abstract
Pseudorandom Number Generators (PRNG) are used in the financial sphere, medicine, game industry, networks and communication, statistical simulation, IT, security, authentication, and cryptography (key management, initialization vectors, one-time passwords). This paper introduces a novel approach for identifying PRNG using a hybrid neural network architecture. The proposed model integrates Recurrent Neural Networks (RNN) and Convolutional Neural Networks (CNN) to enhance the accuracy of classification. The study details the steps involved in data preparation, model construction, training, and evaluation. Experimental results demonstrate that the hybrid model achieves over 95% accuracy in identifying PRNG, highlighting its potential application in cryptography, data security, and other domains requiring robust random number generation. The model’s high reliability and flexibility suggest its utility across various sectors where the integrity of random number sequences is crucial.
Item Type: | Article |
---|---|
Uncontrolled Keywords: | PRNG; source identification; hybrid neural network; RNN; CNN; cryptography; machine learning; classification; data security |
Subjects: | Статті у базах даних > Scopus |
Divisions: | Факультет інформаційних технологій та математики > Кафедра інформаційної та кібернетичної безпеки ім. професора Володимира Бурячка |
Depositing User: | Павло Миколайович Складанний |
Date Deposited: | 06 Dec 2024 09:46 |
Last Modified: | 06 Dec 2024 09:46 |
URI: | https://elibrary.kubg.edu.ua/id/eprint/50187 |
Actions (login required)
![]() |
View Item |