Stackelberg and MAB Models for Decision-Making Process

This paper considers the Stackelberg and the Multi-Armed Bandit (MAB) models in the context of Cognitive Radio (CR) in order to enhance the spectrum learning of the secondary users (SU s), In such context, a SU tries to maximize selfishly its utility function which generally depends on the throughput, transmit power level, and the interference from other users. Recently, MAB model is widely suggested in the literature to learn the availability of channels for a single SU. While, in this work, beside the MAB, we focus on the Stackelberg model to consider more developed scenarios, such as: Learn channels' availability, enhance the Quality of Service, adjust transmit power level and decrease the interference among SUs. We also focus on the priority access, where a leader reacts first and then the followers make their decisions and actions by taking into consideration the action of the leader. We show that the Stackelberg model can represent an optimal solution in a real radio environment to help a SU make a good decision.