While 5G technology offers enormous improvements over previous-generation licensed wireless networks, much of the impetus for pushing 5G into new frequency bands comes from old-fashioned physics, according to experts.
It can be difficult to nail down precisely which frequencies 5G technology will use, in part because it varies heavily across different countries and even among different carriers. There are, however, three main groupings into which most 5G frequencies will be able to fit.
The low end: 650MHz to 1GHz
Lower frequencies – from around 650MHz at the lowest up to around 1GHz – are particularly prized by wireless companies deploying 5G. Signals in this range propagate over relatively long distances, meaning that service providers can cover a huge area with a single access point.
However, there’s a serious limiting factor to 5G use in the lower range of the RF spectrum, according to Gartner analyst Bill Ray.
“[These] are very popular frequencies,” he said. “And the military and TV stations still own a lot of them.”
That’s a problem, because while the channel capacity is the same at the low end as at higher frequencies – that is to say, a 5MHz-wide channel in the 850MHz range offers the same throughput as a 5MHz-wide channel in the 2.6GHz range – the lack of available spectrum means that there simply aren’t enough channels to provide the high connection speeds that 5G advertises.
To attain higher speeds, 5G uses wider channels. “In 3G, the standard transmission was 5MHz wide,” Ray added. “In 5G, we’re talking about slots that are 100MHz wide, so your transmission [channel] can run from 2.4 to 2.5Ghz”
The middle: Sub-6GHz and the spectrum rush
Much of the performance and efficiency increases promised by 5G technology rely on parts of the spectrum that are close to Wi-Fi – the “sub-6GHz” range between 2.4GHz and 6GHz, which is a departure for service providers used to playing slightly lower on the wireless spectrum.
“The sub-6GHz stuff is new spectrum for them,” said Patrick Filkins, a senior research analyst with IDC. “It means bigger channels and latency improvements.”
An additional wrinkle is that there’s free spectrum available in this frequency band, particularly in the citizens broadband radio service/general authorised access (CBRS/GAA) spectrum between 3.5GHz and 3.7GHz. This spectrum uses a system of prioritised access, with incumbents – specifically the U.S.
Navy and satellite ground stations – given first dibs, but others may use the frequencies anywhere where they’re not interfering with those incumbents. Needless to say, the spectrum-hungry carriers are interested.
As individual carriers carve out their own pieces of this valuable spectrum, 5G coverage will continue to expand, and its advantages over Wi-Fi in one particular aspect will become clearer. 5G spectrum is parcelled out to one licensee per geographic area whereas Wi-Fi offers no such exclusivity; anyone can use it anywhere, setting the stage for overlapping signals and interference.
“The way that cellular is deployed is very deterministic – one of the big benefits it has over Wi-Fi is that it’s able to overcome that interference problem that Wi-Fi doesn’t have a great answer for yet,” said Filkins.
The high side: Millimetre wave
For all of its optimisation and sophistication, 5G still requires lots and lots of bandwidth to deliver on its promises of gigabit-scale throughput. And despite the wireless industry’s aggressive pursuit of bandwidth in lower frequencies, there simply isn’t enough available space in the more desirable sub-6GHz and 1GHz ranges.
Enter millimetre-wave technology, which operates at anywhere between 24GHz and 60GHz, depending on which expert you’re talking to and the particular technology involved. These particularly high-frequency bands allow for particularly wide transmission channels, enabling those blazing-fast connection speeds, but there are a host of drawbacks.
“The only good thing about [millimetre-wave] frequencies,” said Ray, “is that they are empty, so there’s plenty of space.”
For one thing, thanks to basic physics, signals in the millimetre-wave range simply don’t propagate very far in comparison to those in lower regions of the RF spectrum, and they don’t penetrate solid objects like walls and windows. This means that, to cover a given area, even within a single building, requires deploying lots and lots of access points.
“It’s not ready for enterprise deployment,” said Filkins. “Right now the story is that it’s not robust, but it has a lot of promise.”
Millimetre-wave technology has been around for at least a decade. The unlicensed wireless world has long had the 802.11ad standard, which is, essentially, Wi-Fi operating at millimetre-wave frequencies. Qualcomm and Huawei both manufacture equipment that operates in that range, but the technology’s limitations, coupled with the continued drought in endpoints that can actually take advantage of it, mean that it has remained on the sidelines.
That won’t be the case forever, of course. Trends in the demand for wireless bandwidth suggest that there will be so much need for spectrum that millimetre-wave will be required to cope with it.
“The rest of the world will use [millimetre-wave] eventually, they’re just not using it now,” said Ray.
For the moment, however – particularly in light of the fact that adding millimetre-wave capable antennas and modems to mobile phones and laptops boosts their per-unit costs by a hefty $20, according to Ray – millimetre-wave will remain a technology of the future.