Global Proportional Fair Scheduling for Networks With Multiple Base Stations

Proportional fair scheduling has attracted much attention and has been well studied in the single-cell network scenario, where only one base station (BS) serves many users. However, for a network where one user may obtain services from multiple BSs geographically located apart, not much work has been done. In this paper, we investigate how to achieve proportional fair allocation of users´ throughput in a wireless network with multiple BSs and multiple users. A low-complexity algorithm is presented. We prove that the developed algorithm is approximately optimal in the sense of maximizing a global proportional fairness utility. Furthermore, two different implementation strategies to achieve the global proportional fair scheduling are proposed: one for networks where inter-BS communications are allowed and the other for networks where only the transmissions between BSs and users are possible. The later implementation makes the algorithm operate in a totally distributed manner. Various simulations confirm that the new global fair scheduling scheme outperforms the single-cell proportional fair scheme in terms of the average user throughput, user throughput oscillation, and fairness. Low-complexity and distributed implementation achieves almost the same performance as the centralized counterpart.