Innovative Applications of Nature-Inspired Algorithms in Cryptographic Communication

Abstract

The increasing complexity of modern communication systems poses significant challenges to traditional cryptographic methods, especially in dynamic, resource-constrained, and adversarial environments. Traditional encryption methods frequently face challenges in achieving an optimal balance between key unpredictability, computational efficiency, and resistance to adaptive attacks. To address these challenges, we propose HSIO-Crypto, a novel cryptographic framework that integrates Genetic Algorithms (GA) and the Remora Optimization Algorithm (ROA) for adaptive key generation and encryption scheduling. The method dynamically evolves encryption keys with high entropy and injects stochastic variation into the encryption process to resist structural and entropy-based attacks. Extensive experiments were conducted across four benchmark datasets (Wireless, Elliptic Curve exchanges (ECC), KDD99, and IoT), with comprehensive comparisons against six state-of-the-art baselines. Results demonstrate that HSIO-Crypto achieves a 17.3% improvement in key strength, a 22.8% reduction in encryption latency, and a 14.6% increase in throughput over GA-Cryptanalysis, while maintaining strong deployment performance on IoT-class devices. Additionally, the proposed method demonstrates enhanced resistance to adversarial inference attacks, achieving an average robustness score 12.5% higher than the strongest baseline. The results show that bio-inspired optimizers can make encryption both fast and secure, even on low-power devices in unstable networks. HSIO-Crypto proves this by letting the cipher adjust itself in real time.