From: Q-learning-based dynamic joint control of interference and transmission opportunities for cognitive radio