Compensation-Based Game for Spectrum Sharing in the Gaussian Interference Channel

This paper considers an optimization problem of sum-rate in the Gaussian frequency-selective channel. We construct a competitive game with an asymptotically optimal compensation to approximate the optimization problem of sum-rate. Once the game achieves the Nash equilibrium, all users in the game will operate at the optimal sum-rate boundary. The contributions of this paper are twofold. On the one hand, a distributed power allocation algorithm called iterative multiple waterlevels waterfilling algorithm is proposed to efficiently achieve the Nash equilibrium of the game. On the other hand, we derive some sufficient conditions on the convergence of iterative multiple waterlevels water-filling algorithm in this paper. Through simulation, the proposed algorithm has a significant improvement of the performance over iterative water filling algorithm and achieves the close-to-optimal performance.