Empirical study set to help industry start exploiting 5G deployed in the mmWave bands
CRNA researchers and partners have tested 5G mmWave performance in different scenarios. In this blog post Research Scientist Foivos Michelinakis presents the results from the study.
The introduction of the millimetre Wave (mmWave) frequency bands greatly enhances 5G performance compared to previous mobile communications technologies such as 4G-LTE and is one of the cornerstones of future generations of mobile networks. It enables 5G to provide much higher bandwidth with significantly less delay than 4G, which in turn will allow industries to develop their very own bespoke applications.
But there are also challenges that need to be dealt with. It’s a well-known fact that mmWaves are highly sensitive to the environment in which they are deployed. Any measurements done in a lab may therefore be invalid when applied outside. CRNA researchers and their partners at Telenor Research and University of Padova have tested the mmWave performance in a few different scenarios in order to give different industries a baseline of what performance they can expect from 5G, for their specific use case. After all, it has to work out there in the real world, not only in laboratories. As far as we know, these tests are the first of their kind performed on a commercial mmWave Base Station.
Some example scenarios were:
Water: What happens when mmWaves travel above water? This could be highly relevant for instance for aquaculture and fish farms.
Dense vegetation: How do the mmWaves cope in forests and areas with dense vegetation?
Urban environment: And how do they cope when the signal is blocked by buildings, moving objects or moving humans?
Some measurement locations: 1) close to water, 2) 6 meters above water, 3) Line of Sight, 4) rain, 5) foliage blockage.
The Base Station uses a technique called beamforming, which directs signals towards the location of a specific device. These directed signals are called “beams” and our Base Station uses 16 of them. When there is no obstacle between the user device and the Base Station, a condition known as Line-of-Sight (LoS), we experience the strongest signal and thus the best performance achieving speeds higher than 1Gbps. Our measurements confirmed what is already observed in the literature: the range of mmWave cells is strongly affected by the presence of obstacles, such as buildings, trees, or human bodies. Despite that, even when we are not able to achieve LoS, a series of phenomena: signal diffraction, reflections from surrounding objects, and multipath propagation may compensate to a degree the signal loss and achieve speeds in the range of several hundreds Mbps, when the distance to the Base Station is small.
Neighbouring beams have slightly overlapped coverage regions. This is good for robustness (body blockage, moving cars, etc), since there is a higher chance of having at least a good beam at any time. The other benefit of this slight overlap (or closely spaced beams) is to have a smooth user experience as a user moves from one beam’s coverage region to another beam's.
For what concerns the effect of foliage, we observed a degradation of the signal quality when the LoS is obstructed by trees, particularly when the leaves and branches moved by (even light) wind. The measurements have also revealed that wide water surfaces, especially in presence of waves, can generate time-varying scattering phenomena that affect the stability of the received signal. This is pronounced if the receiver is higher than the water surface and thus collects more water-reflected waves. The propagation of mmWaves on water surfaces is, hence, critical and would require further investigation to determine the limitations of links involving floating stations.
Finally, we compared our received signal power measurements with simulations based on empirical models to check how accurately statistical models can predict the received signal at different locations. We observe big deviations between our measurements and the simulations. These deviations are pronounced in scenarios without LoS. The complexity of the propagation environment, makes impossible the use of accurate channel models. This shows the necessity of measurements, especially in complex non LoS environments, in order to predict the cell coverage and recognize coverage holes.
The above observations should help interested stakeholders make more informed decisions when deploying 5G solutions utilising the mmWave spectrum, providing a baseline for further work.
The article Sectors, Beams and Environmental Impact on Commercial 5G mmWave Cell Coverage: an Empirical Study has been written by Salman Mohebi (University of Padova ) Foivos Michelinakis (SimulaMet), Ahmed Elmokashfi (SimulaMet), Ole Grøndalen (Telenor Research), Kashif Mahmood (Telenor Research), Andrea Zanella (University of Padova)
The full article has been submitted for peer review. A draft version of the article can be found here.