Bayesian game-based user behavior analysis for spectrum mobility in cognitive radios

Abstract

In cognitive radios (CRs) secondary users (SUs) opportunistically exploit the spectrum, vacate the channels detecting primary user (PU) activity and cause spectrum mobility. This spectrum mobility is inevitable to truly benefit from CRs however, there exists a cost-benefit tradeoff every time an SU wants to switch its channel. For SUs to achieve this tradeoff, they must take an optimum switching decision and hence it is mandatory to analyze the decision-making behavior of SUs such that when and how should they switch the channels. This behavior is inescapable to study if CR users have to behave rationally. In this paper, we investigate the mobility-incurred effective decision making of an SU, given the trade-off between cost and benefit. The problem is formulated as Bayesian game with finite spaces and continuous types in two scenarios; with two-players and with n-players. The solution concept of this game provides multiple equilibrium, hence the existence of both pure NE (Nash Equilibrium) and mixed BNE (Bayesian Nash Equilibrium) for both scenarios are proved. In addition, the randomization nature of the players has been analyzed and it has been proved that the behavior of the SU is rational and it tends to maximize its payoff.