Assessing the Performance of a 60-GHz Dense Small-Cell Network Deployment from Ray-Based Simulations

TL;DR

This paper evaluates the performance of a 60-GHz dense small-cell network in a suburban environment using ray-based simulations, focusing on deployment, propagation, and interference considerations for 5G mmWave systems.

Contribution

It introduces an enhanced ray-tracing simulation method incorporating detailed vegetation modeling to analyze 60-GHz small-cell network deployment in realistic environments.

Findings

Ray-tracing effectively captures channel properties affecting beam-steering.

Inter-cell interference varies with user density and obstructions.

Network performance is sensitive to local environmental factors.

Abstract

Future dense small-cell networks are one key 5G candidates to offer outdoor high access data rates, especially in millimeter wave (mmWave) frequency bands. At those frequencies, the free space propagation loss and shadowing (from buildings, vegetation or any kind of obstacles) are far stronger than in the traditional radio cellular spectrum. Therefore, the cell range is expected to be limited to 50 - 100 meters, and directive high gain antennas are required at least for the base stations. This paper investigates the kind of topology that is required to serve a suburban area with a small-cell network operating at 60 GHz and equipped with beam-steering antennas. A real environment is considered to introduce practical deployment and propagation constraints. The analysis relies on Monte-Carlo system simulations with non-full buffer, and ray-based predictions. The ray-tracing techniques are today identified as a relevant solution to capture the main channel properties impacting the beam-steering performance (angular dispersion, inter-link correlation); and the one involved in the present study was specifically enhanced to deal with detailed vegetation modeling. In addition to the user outage, the paper evaluates the evolution of the inter-cell interference along with the user density, and investigates the network behavior in case of local strong obstructions.