A graph-based interference topology control for ultra-dense networks

In densely deployed cellular networks, we can proactively avoid interference by controlling interference topology rather than passively mitigating the interference after it has been generated. While there are many ways to control the interference topology, we employ resource allocation as an example to illustrate its performance gain. Traditional graph-based resource allocation methods are often developed from the minimal coloring problem, which can improve the performance of cell-edge users, but suffer from severe loss of network throughput. In this paper, we propose a graph-based resource allocation method to maximize the resource usage efficiency (RUE) under interference-free constraint, which is related to the maximal independent sets of the interference topology graph. Considering that the Max RUE method is inclined to allocate almost all resources to a few cells, causing unfairness among the users, we further propose another graph-based resource allocation method to achieve a good trade-off between the RUE and user fairness. Simulation results show that the cell-edge throughput is significantly boosted with minor degradation in average throughput of the network.