A Game Theoretic Approach to Spectrum Pricing in Two-Tier HetNets with Incomplete Information

This paper studies a scenario as an oligopolistic game in an cognitive radio market where primary networks compete with each other to obtain a maximum profit by selling unused spectrum. Assuming that one of the primary networks does not provide information about the amount of available spectrum for the auction, a Bertrand game with incomplete information is used as a model for this market. A Nash Bayesian Game is proposed as a solution, which achieves an equilibrium for the profit earned by both primary networks. In this game the cost, demand and profit are functions of spectral efficiency and spectrum substitutability, which enables us to analyse their impact on the market. As a result of this work we conclude that a secondary network will choose the cheapest spectrum price in the market regardless the QoS offered by a primary network. However when the market is very competitive there is an uncertain behaviour in the network due the lack of information provided for one of the primary networks. This is measured as a 1.4dB quality of service range where it is not certain that the lowest price will win.