Performance evaluation of dynamic interference coordination algorithms

Dynamic interference coordination has been widely studied and the core coordination algorithm is the keypoint to determine its performance. However, there lacks a complete performance evaluation of existing coordination algorithms in the literature. This paper focuses on average utility, fairness, and time complexity of various algorithms in homogeneous and heterogeneous 2-cell scenarios. We find that greedy and alternated horse racing achieve both high utility and fairness. Alternated inverse quick pairing reaches high utility and fairness when inter-cell interference (ICI) is not serious. Alternated quick pairing, alternated inverse quick pairing, and alternated horse racing out-perform others in terms of time complexity. To sum up, we conclude that, for the case with serious ICI, alternated horse racing should be promising, while for the case when ICI is not quite serious, alternated inverse quick pairing becomes probably the best choice.