Sequential Bargaining in Cooperative Spectrum Sharing: Incomplete Information with Reputation Effect

Cooperative spectrum sharing can effectively improve spectrum usage by allowing secondary users (SUs) to dynamically share the licensed bands with primary users (PUs). Meanwhile, an SU can relay a PU´s traffic to improve the PU´s effective data rate. In this paper, we consider a sequential spectrum bargaining process to achieve cooperative spectrum sharing between one PU and one SU over multiple time slots. The SU may be a Low type or a High type, depending on its energy cost. Such information is private to the SU and is unknown to the PU. We model such a dynamic bargaining with incomplete information as a dynamic Bayesian game, and characterize several types of equilibria under different system parameters. In particular, we show that a Low type SU may maximize its total utility by utilizing the reputation effect, i.e., rejects profitable offers initially in order to create the reputation of a High type SU.