QoS-aware Scheduling in 5G Wireless Base Stations

Abstract: 5G and beyond networks are expected to support flows with varied \emph{Quality-of-Service (QoS)} requirements under unpredictable traffic conditions. Consequently, designing policies ensuring optimal system utilization in such networks is challenging. Given this, we formulate a long-term time-averaged scheduling problem that minimizes a weighted function of packets dropped by the 5G wireless base station. We then present two policies for this problem. The first is a delay-guaranteed near-optimal policy, and the second is a delay-guaranteed sub-optimal policy that provides flow isolation. We perform extensive simulations to understand the performance of these policies. Further, we study these policies in the presence of a closed-loop flow rate-control mechanism.