Work

State-Based Potential Game Approach for Anti-Jamming Channel Access via Reinforcement Learning