Modeling Non-Uniform UE Distributions in Downlink Cellular Networks

TL;DR

This paper introduces a new method to model and analyze cellular networks with non-uniform user distributions by conditionally thinning base station processes, enabling separate analysis of cell interior and edge UEs.

Contribution

A novel tractable sampling framework using conditional thinning of BS processes to study non-uniform UE distributions in cellular networks.

Findings

Enables separate analysis of cell interior and edge UEs.

Provides a generative model for non-uniform UE distributions.

Facilitates analysis of deployment scenarios with dependent UE and BS locations.

Abstract

A recent way to model and analyze downlink cellular networks is by using random spatial models. Assuming user equipment (UE) distribution to be uniform, the analysis is performed at a typical UE located at the origin. While this method of sampling UEs provides statistics averaged over the UE locations, it is not possible to sample cell interior and cell edge UEs separately. This complicates the problem of analyzing deployment scenarios involving non-uniform distribution of UEs, especially when the locations of the UEs and the base stations (BSs) are dependent. To facilitate this separation, we propose a new tractable method of sampling UEs by conditionally thinning the BS point process and show that the resulting framework can be used as a tractable generative model to study cellular networks with non-uniform UE distribution.