SIR distribution and scheduling gain of normalized SNR scheduling in Poisson networks

Abstract

In most stochastic geometry analyses of cellular networks, a receiving user is selected randomly. Thus, the network performance is equivalent to that obtained with the use of a round-robin scheduler; in other words, channel-aware user scheduling is not applied. The first objective of this study is to clarify the type of channel-aware user scheduling that can be utilized with stochastic geometry. In both Poisson bipolar networks and Poisson cellular networks, the signal-to-interference power ratio (SIR) distribution, average data rate, and scheduling gain of a normalized SNR scheduler are derived. The scheduler selects the user with the largest SNR normalized by the short-term average SNR, and it is equivalent to the well-known proportional fair scheduler when the data rate is proportional to the SNR. The second objective is to discuss the achievable scheduling gain with stochastic geometry when compared with round-robin scheduling. The value of the scheduling gain is examined using numerical integration results.