Multi-Channel Transmission Scheduling Based on Reinforcement Learning in Cyber-Physical Systems Under DoS Attacks

This paper advances a strategic paradigm for sensor power scheduling within Cyber- Physical Systems (CPSs) under adversarial scenarios, leveraging a Markov Stackelberg game combined with a Signal-to-Interference-plus-Noise Ratio (SINR) framework. A novel approach, which introduces reinforcement learning for the adaptive optimization of power distribution, is proposed to ascertain energy efficient sensor transmission tactics in the presence of jamming. Empirical simulations are given to validate the superiority of the proposed algorithm, underscoring its effectiveness in diminishing estimation inaccuracies and alleviating the adversarial influence.