Novel MIMO Packet-Based Proportional Fairness Scheduling Framework

The key issue we address in this paper concerns how the available bandwidth within the downlink of a multiple-input multiple-output (MIMO) cellular network should be shared between competing users to achieve both fairness and high system throughput. We propose a high performance packet scheduling framework, called MIMO packet-based proportional fairness (MP-PF). The scheduler is work conserving and based on a novel proportional fairness concept that takes the packet length, users' backlogs, and users' guarantees into consideration. Furthermore, we gain an efficient packet-based framework in the MAC layer. Intensive simulation studies that took into consideration the traffic characteristics and the mobility of the users were used to demonstrate the superior performance of our scheduling framework for both the average delay of users and the time/service fairness. We also propose two useful equations for the time/service comparison of MIMO schedulers.