Hard deadline constrained multiuser scheduling for random arrivals

In this work, an opportunistic scheduling scheme for a large multiuser system is proposed. A group of users with good channels are scheduled simultaneously for data transmission and separated by means of superposition coding. The proposed scheduling scheme is analyzed in the large system limit. Random packet arrivals are modeled as constant arrivals with random content size. Transmission thresholds are optimized in such a way that the system energy is minimized while obeying a strict upper bound on the packet delay. We find that the state space representations of systems with either constant or random arrivals are equivalent. Thus, the thresholds optimized for constant arrivals in earlier work are valid for systems with random arrivals as well. Furthermore, we address the option of intentional packet dropping and the trade-off between packet drop rate and required system energy.