Long-term proportional fairness over multiple cells

In current cellular networks, Proportional Fair Scheduling (PFS) is performed separately at each base station. Such single-cell scheduling is not aware of the data rate that mobile users received in previously traversed cells. In this paper, we will show that, by lacking this information, Single-cell PFS fails to provide Proportional Fairness (PF) in the long-term as mobile users move over multiple cells. To overcome this limitation, we extend Single-cell PFS by the long-term average rate a mobile user received in previous cells. By being aware of this scheduling history, our Multi-cell PFS approach achieves substantially higher fairness and throughput compared to the traditional single-cell PFS approach.