摘要

In this paper, we study the spectrum sharing problem in the cognitive network where two secondary users coexist and interfere with each other in an unlicensed channel. Each user in our scenario is independent and is unaware of the characteristic of its rival but its statistics (imperfect information). An important issue is the interaction among selfish independent wireless communication systems (secondary users) who aim to maximize their own current utility. We provide a self-enforcing spectrum sharing framework, which has the advantage of not requiring a control center or communicating between two users. The short-term profits of users gained before cooperation being builded are asymmetrical. The long-term profit is used as the user's utility. Users improve their long-term profits by creating mutual cooperation actively. The historical action of the user is used by its rival to evaluate/update the current action. The action updating algorithm is provided. Moreover, we investigate the stability of the cooperation. As demonstrated, the proposed scheme can build mutual cooperation through extensive simulations. The existence of equilibrium is confirmed by the repeated game. And our scheme can achieve a good trade-off between short-term profits and long-term profits.

全文