Analytical Model of Proportional Fair Scheduling in Interference-Limited OFDMA/LTE Networks

Various system tasks like interference coordination, handover decisions, admission control etc. in upcoming cellular networks require precise mid-term (spanning over a few seconds) performance models. Due to channel-dependent scheduling at the base station, these performance models are not simple to obtain. Furthermore, upcoming cellular systems will be interference-limited, hence, the way interference is modeled is crucial for the accuracy. In this paper we present an analytical model for the SINR distribution of the scheduled subcarriers of an OFDMA system with proportional fair scheduling. The model takes the precise SINR distribution into account. We furthermore refine our model with respect to uniform modulation and coding, as applied in LTE networks. The derived models are validated by means of simulations. In additon, we show that our models are approximate estimators for the performance of rate-based proportional fair scheduling, while they outperform some simpler prediction models from related work significantly.

I. Introduction

Cellular networks are facing a considerable increase in mobile data traffic over the next ten years [1]. The current traffic impact generated from smartphones is to some extent evidence for the development still ahead. To satisfy these demands, ambitious requirements are set for the upcoming 4th generation of cellular networks like IEEE 802.16 [2] and 3GPP's LTE (long term evolution) [3]. The demand of higher capacity on the one hand and the presence of scarce frequency resources on the other, poses challenging aspects on their system design. Frequency reuse of one is applied for a flexible usage of the resource allocation especially for asymmetrical load distributions. Additionally, feedback based resource allocation schemes (scheduling) are introduced for an efficient exploitation of the available resources.