Approximation of Uplink Inter-Cell Interference in FDMA Small Cell Networks

TL;DR

This paper proves that uplink inter-cell interference in FDMA small cell networks can be approximated by a lognormal distribution under certain conditions, simplifying performance analysis for dense 5G deployments.

Contribution

It provides the first analytical proof of the lognormal approximation for uplink interference in small cell networks and identifies conditions independent of user equipment distribution shapes.

Findings

Lognormal approximation applies under low ratio of third absolute moment to variance.

The condition is satisfied in dense/ultra-dense SCNs, aiding 5G network analysis.

Results are applicable beyond Poisson deployment models.

Abstract

In this paper, for the first time, we analytically prove that the uplink (UL) inter-cell interference in frequency division multiple access (FDMA) small cell networks (SCNs) can be well approximated by a lognormal distribution under a certain condition. The lognormal approximation is vital because it allows tractable network performance analysis with closed-form expressions. The derived condition, under which the lognormal approximation applies, does not pose particular requirements on the shapes/sizes of user equipment (UE) distribution areas as in previous works. Instead, our results show that if a path loss related random variable (RV) associated with the UE distribution area, has a low ratio of the 3rd absolute moment to the variance, the lognormal approximation will hold. Analytical and simulation results show that the derived condition can be readily satisfied in future dense/ultra-dense SCNs, indicating that our conclusions are very useful for network performance analysis of the 5th generation (5G) systems with more general cell deployment beyond the widely used Poisson deployment.