Pseudo Random Number Generator Based on Recurrent Neural Networks

View at publisher Summarise with VIRAT
Authors: Prince Prasad, R. Amutha, Vignesh M
Cited by: 0
References: 15

Summary

No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.

Summarise with VIRAT Free to try with your Science Catalogs account.

Abstract

This paper proposes a Pseudo Random number generator (PRNG) using recurrent neural network. The author investigated the application of Artificial Neural Networks (ANNs) for statistically random sequence generation, utilizing their capacity for learning intricate temporal patterns. Two architectures namely, Long Short-Term Memory (LSTM) and Gated Recurrent Unit (GRU) were trained on chaotic time series data from the logistic map. The performance of each model was tested using entropy, correlation coefficients, and autocorrelation/cross-correlation measures, with randomness being confirmed by the NIST Statistical Test Suite. Both LSTM and GRU exhibited the best prediction abilities. The results pave the way for further research in the area of neural cryptographic primitives and demonstrate how deep learning can be utilized to enhance randomness generation for secure systems.

Fields of Study & Subjects

Chaos-based Image/Signal Encryption Neural Networks and Applications Cellular Automata and Applications

Publication details

Journal / source
International Conference on Signal Processing and Electronics Design (ICSPED)
Pages
157–163
Publication date
23 January 2026
Language
EN