Intelligent Frequency-Hopping Strategies for Securing the Physical Layer Based on Learning for the Stochastic Dispersion Problem

This article addresses the security of the physical layer against adaptive jamming attacks in wireless networks. We propose an intelligent frequency-hopping strategy that combines the Q-learning algorithm with a stochastic spreading mechanism to optimize transmission resilience. Unlike conventional FHSS, our approach allows the transmitter to autonomously learn compromised channels and adjust its hopping policy in real time. Stochastic spreading ensures maximum signal unpredictability, preventing any prediction by a cognitive adversary. Simulations demonstrate rapid convergence of the algorithm and a significant increase in the transmission success rate, reaching 100% at an SNR of 8 dB. Complexity analysis confirms that the induced latency is compatible with real-time requirements. This work paves the way for self-organizing and robust communication systems for future 5G/6G networks and the Internet of Things.