Distributed ABS-Slot Access in Dense Heterogeneous Networks: A Potential Game Approach With Generalized Interference Model

This paper investigates the problem of almost blank subframe (ABS)-slot access in the heterogeneous networks to optimize network performance under the ABS scheme. Since maximizing network throughput requires global information and is not feasible in large scale or dense networks, the problem is solved from an interference minimization perspective after investigating the inherent relationship between local interference and achievable throughput. Since traditional binary interference model does not consider accumulated interference and thus results in inaccuracy, we take the accumulated interference into account and propose a generalized interference model, in which small cells are classified into different distant neighborhoods. An interference minimization game is formulated and is then proved to be an exact potential game, which has at least one pure Nash equilibrium (NE), and the best pure strategy NE point is a global optimum of minimizing aggregate interference. Combining the relationship between throughput and interference, the throughput maximization problem is proved to be an ordinal potential game, which possesses nice properties as well. A distributed algorithm is proposed to reach the NE, which minimizes the aggregate interference and maximizes the network throughput globally or locally. Furthermore, simulation results show that the proposed generalized interference model outperforms the traditional binary interference model.