A Game theoretical approach for spectrum sharing in cognitive radio systems with payoff perturbations

We study a spectrum allocation game where K cognitive radio systems compete for a number of free resource blocks in frequency domain. In particular, we concentrate on the impact of the perturbations in the payoff functions and the problem of equilibrium selection. Based on the simulated data throughput, we first address the problem of designing an appropriate spectrum allocation game. Our study shows that if we can persistently increase the accuracy of the simulation results, we are able to find the approximate Nash equilibria, which are arbitrarily close to the Nash equilibria of the real spectrum allocation game. Then, we study the problem of equilibrium selection, where we exploit the concepts of payoff dominance and the focal-point effect to select a unique Nash equilibrium. At last, we demonstrate the performance of the spectrum allocation game in an LTE system context.