Survey on AI-Enhanced Chaotic Key Generation for Public Key Cryptography

Abstract

According to recent trends, integrating chaotic systems with artificial intelligence (AI) may be a promising way to improve cryptographic key generation, especially in public key cryptography.. This survey paper provides the tools used in chaos-based key generation augmented by AI. It includes recent technological developments, the basic ideas behind these approaches and real-world applications. Because they produce extremely random and difficult-to-guess keys, other AI models like neural networks and reinforcement learning improve the functionality of chaotic maps.. Other contemporary issues discussed in the paper include the real-time use of the methods, clarity in their decisions, quantum attack protection, and standards. In short, employing AI to enhance chaotic systems sounds like a powerful way to produce public key cryptography, but there continue to be some significant challenges to solve, like potentiality the system is simple to understand, guarantee it creates well, and master it prepared for genuine use.

Introduction

The paper explores how Artificial Intelligence (AI) can enhance chaotic systems for generating highly secure cryptographic keys in Public Key Cryptography (PKC). It highlights the growing need for secure digital communication, especially in the face of modern threats, and evaluates the potential, benefits, limitations, and future directions of integrating AI with chaos theory.

1. Background Concepts

• Public Key Cryptography (PKC):

  • Uses a public key for encryption and a private key for decryption.

  • Common algorithms: RSA, Elliptic Curve Cryptography (ECC), Lattice-based cryptography (quantum-safe).

• Chaotic Systems in Cryptography:

  • Use chaotic maps (e.g., Logistic Map, Lorenz System) to generate pseudo-random keys.

  • Offer high sensitivity and complexity, but can be unstable or inefficient.

• Role of AI:

  • AI techniques like machine learning, neural networks, genetic algorithms (GA), and particle swarm optimization (PSO) are used to:

    • Improve the randomness of chaotic sequences.

    • Enhance system performance and adaptability.

    • Optimize chaotic parameters for better security.

2. Literature Review Highlights

• Researchers have combined chaos with AI to improve:

  • Key randomness.

  • Resistance to attacks.

  • Efficiency in low-resource environments.

• Examples include:

  • Use of neural networks (TPMs) and LSTM/GRU models.

  • Image encryption using chaotic maps.

  • GA and PSO for optimizing ECC key generation.

3. Comparative Analysis of Key Generation Techniques

Key metrics compared:

• Entropy (randomness)

• NPCR/UACI (sensitivity for image encryption)

• Key Space Size (number of possible keys)

• Attack Resistance

• Computational Efficiency

• Suitability for PKC and post-quantum systems

Key Findings:

• Older methods (e.g., Pareek 2006) lack post-quantum readiness.

• Newer models (e.g., Song 2025, Bouke 2025) offer:

  • Maximum entropy

  • Post-quantum security

  • High attack resistance

  • Excellent PKC suitability

4. Applications

AI-chaos-based key generation is applicable in:

• Secure communications

• Internet of Things (IoT)

• Cloud data security

• Blockchain systems

• Biometric authentication

• Post-quantum cryptography

5. Challenges & Open Problems

Despite progress, several issues remain:

• Adversarial AI attacks reducing randomness.

• High computational costs not suited for IoT/edge devices.

• Lack of interoperability with current PKC systems.

• Training privacy risks, revealing generator structures.

• Inflexibility in dynamic key adaptation.

• Decentralized AI training adds synchronization complexity.

• Post-quantum compatibility not fully resolved.

• Ethical concerns around AI transparency, autonomy, and governance.

Conclusion

The idea of AI-enhanced chaotic key generation is using the inherent randomness in chaotic systems, specifically, Logistic, Henon, and Lorenz maps, in concert with AI optimization methods to generate highly complex, dynamically complicated, and arguably secure keys workable within IoT or edge computing environments, and even resource-constrained systems. There are also still major challenges such as standardized evaluation metrics, good performance in the real world, energy efficiency, interpretability of AI models, adaptive key evolution, interoperation with existing cryptography systems, post-quantum applicability, privacy-preserving training, and governance, and they are a major research focus required to make AI deployable in practice at scale in secure systems.

References

[1] Diffie, W., Hellman, M. E., & Ellis, J. (1976, June). Public key cryptography. In IEEE International Symposium on Information Theory. [2] Kocarev, L., & Lian, S. (Eds.). (2011). Chaos-based cryptography: Theory, algorithms and applications (Vol. 354). Springer Science & Business Media. [3] Hafezi, R. (2019). How Artificial Intelligence Can Improve Understanding in Challenging Chaotic Environments. World Futures Review, 12(2), 219-228. https://doi.org/10.1177/1946756719880539 (Original work published 2020) [4] Menon, Unnikrishnan, Anirudh Rajiv Menon, and Atharva Hudlikar. \"A novel chaotic system for text encryption optimized with genetic algorithms.\" arXiv preprint arXiv:2011.00575 (2020). [5] Manasi Hazarika, 2014, A Review of Chaos-Based Image Encryption Techniques, INTERNATIONAL JOURNAL OF ENGINEERING RESEARCH & TECHNOLOGY (IJERT) Volume 03, Issue 02 (February 2014), [6] Zhang, L. Y., Hu, X., Liu, Y., Wong, K. W., & Gan, J. (2014). A chaotic image encryption scheme owning temp-value feedback. Communications in Nonlinear Science and Numerical Simulation, 19(10), 3653-3659. [7] Kumar, S., & Sharma, D. (2023). Key Generation in Cryptography Using Elliptic-Curve Cryptography and Genetic Algorithm. Engineering Proceedings, 59(1), 59. https://doi.org/10.3390/engproc2023059059 [8] Alibrahim, H., & Ludwig, S. A. (2021, December). Image encryption algorithm based on particle swarm optimization and the chaos logistic map. In 2021 IEEE Symposium Series on Computational Intelligence (SSCI) (pp. 01-08). IEEE. [9] Sarkar, A. (2021). Secure exchange of information using artificial intelligence and chaotic system-guided neural synchronization. Multimedia Tools and Applications, 80(12), 18211-18241. [10] Kadir, A., Azzaz, M. S., & Kaibou, R. (2023, March). Chaos-based key generator using artificial neural network models. In 2023 International Conference on Advances in Electronics, Control and Communication Systems (ICAECCS) (pp. 1-5). IEEE. [11] Mahdi, A. A., & Hoobi, M. M. (2025). Robust and Efficient Methods for Key Generation using Chaotic Maps and A2C Algorithms. Mesopotamian Journal of CyberSecurity, 5(1), 301-318. [12] Song, K., Imran, N., Chen, J. Y., & Dobbins, A. C. (2025). A Hybrid Chaos-Based Cryptographic Framework for Post-Quantum Secure Communications. arXiv preprint arXiv:2504.08618. [13] Bouke, M. A. (2025). The Hashed Fractal Key Recovery (HFKR) Problem: From Symbolic Path Inversion to Post-Quantum Cryptographic Keys. arXiv preprint arXiv:2506.04383. [14] Paul, B. (2023). A Novel Low-Power Encryption Scheme Based on Chaotic Dynamic Triple Pendulum System for Wide Range of Applications. Authorea Preprints. [15] Banu, Y., Rath, B. K., & Gountia, D. D. (2025). Analyzing cryptographic algorithm efficiency within graph-based encryption models. Frontiers in Computer Science, 7, 1630222.

Copyright

Copyright © 2025 Shajeela Beegom B A, Dr. Karthigai S, Dr. Selvanayaki K. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.