A Potential Game Approach Towards Distributive Interference Management in OFDMA-Based Femtocell Networks

Abstract

This paper proposes a distributive interference management for OFDMA-based hierarchical macro/femto-cell networks. Specifically, the joint resource block (RB) and power allocation of the downlink case is modeled by the potential game (PG) approach, which is proved to exhibit pure and unique Nash Equilibrium (NE). In the designed PG, the femto base stations (HBSs) are the players, and the RB and power allocation are the set of actions that each player can perform. The utility function of the game is designed in a way such that it guarantees the existence of femto cells within macro cell. To this end, the utility function incorporates all the sources of interference, which include co-tier (femto-femto) and cross-tier (femto-macro), and also the reward of each player. The game is solved by exploiting the best response dynamics, which corresponds to the selection of RBs and power allocation by HBSs. In addition, particle swarm optimization is taken into account for the joint allocation task for each iteration of the game. Simulation results are presented in terms of the performance metrics, which include convergence, impact of varying RBs and fairness. For the sake of comparison, RB random allocation and non-cooperative power optimization game is considered.