Performance Limits of Network Densification

TL;DR

This paper analyzes the fundamental performance limits of network densification in 5G ultra-dense networks, revealing three distinct scaling regimes for user performance and offering insights for optimal deployment strategies.

Contribution

It introduces a unified framework incorporating pathloss and fading effects, identifying three performance scaling regimes and providing qualitative comparison tools for transmission techniques.

Findings

User performance can increase, saturate, or decay with network density.

Coverage density and area spectral efficiency can scale better than user performance.

Simulation results verify the theoretical scaling regimes.

Abstract

Network densification is a promising cellular deployment technique that leverages spatial reuse to enhance coverage and throughput. Recent work has identified that at some point ultra-densification will no longer be able to deliver significant throughput gains. In this paper, we provide a unified treatment of the performance limits of network densification. We develop a general framework, which incorporates multi-slope pathloss and the entire space of shadowing and small scale fading distributions, under strongest cell association in a Poisson field of interferers. First, our results show that there are three scaling regimes for the downlink signal-to-interference-plus-noise ratio (SINR), coverage probability, and average per-user rate. Specifically, depending on the near-field pathloss and the fading distribution, the user performance of 5G ultra dense networks (UDNs) would either monotonically increase, saturate, or decay with increasing network density. Second, we show that network performance in terms of coverage density and area spectral efficiency can scale with the network density better than the user performance does. Furthermore, we provide ordering results for both coverage and average rate as a means to qualitatively compare different transmission techniques that may exhibit the same performance scaling. Our results, which are verified by simulations, provide succinct insights and valuable design guidelines for the deployment of 5G UDNs.