A Pricing-Based Spectrum Leasing Framework with Adaptive Distributed Learning for Cognitive Radio Networks

In this paper, we consider the decentralized scenario of spectrum leasing, whereby a primary user (PU) who owns the spectrum resource, may lease a part of her licensed spectrum to a secondary (SU). We propose a pricing-based spectrum leasing framework between the PU and the SU. The spectrum leasing problem can be depicted by a non-cooperative game where: on one hand, the PU plays the seller and attempts to maximize its own utility by setting the price of spectrum. On the other hand, the SU (i.e., the buyer) has to decide whether to accept the leasing offer or to decline it, while seeking to maximize her own utility. Next, we characterize the Nash equilibria of the induced game for both pure strategies and mixed mixed strategies. We also propose numerous learning algorithms that allow cognitive users to learn their optimal strategies and payoffs for both continuous and discontinuous actions. Simulation results evaluate our model and show the behaviour (accuracy and speed of convergence) of the proposed learning algorithms.