Performance analysis of a Stackelberg game-theoretical strategy design in cognitive radio networks

We investigate the strategic behavior and analyze the performance of multiple secondary users (SUs) in a hierarchical dynamic spectrum sharing (DSS) with asymmetric information structure among them into consideration. This hierarchical DSS can be well captured using the Stackelberg capacity maximization game (SCMG) model, wherein Stackelberg equilibrium solution (SES), a potentially more effective equilibrium solution, is investigated. More importantly, we conclude that SES is not always better than the Nash equilibrium solution (NES), further the mild and specified conditions to guarantee this optimality are provided and multiple SUs prefer to be followers in SCMG. Meanwhile, closed-form power selection solutions are derived for the leader and follower players, respectively, based on which distributed algorithms in similar with iterative water-filling algorithm are designed to approximate the final SES with low implementation complexity and limited information exchange. Simulation results further verify the performance analysis and dynamic behaviors.