You've seen the shouty adverts and breathless tech articles. 5G is fast and it's here and you should drop everything and buy a 5G phone right now.
EE is the first UK network to offer 5G plans. You can pre-order phones attached to them at EE’s website now, starting at £54 per month for the Oppo Reno 5G. The price rises to £69 if you want a phone your friends have actually heard of, like the Samsung Galaxy S10 5G, or data allowances that justify faster internet. Vodafone plans to follow suit more-or-less imminently, in July. Three says its 5G network is coming later this year.
The first 5G phones are here too: the OnePlus 7 Pro 5G, Samsung Galaxy S10 5G, Oppo Reno 5G, Xiaomi Mi Mix 3 5G and LG V50 5G. Each is either very similar to a top-end 4G phone, or has tweaks like a larger screen to help justify the often-significant added cost.
Samsung’s Galaxy S10 5G has a 6.7-inch screen, four cameras on its back (rather than three) and two on the front. It costs a packet and has the tech to match. The Oppo Reno 5G costs significantly less - Oppo is relatively new to the UK, but is popular in China and India.
5G is firing on all cylinders. And if you come across the right articles online, or find the right circles on Twitter, you’ll see photos of handfuls of technology journalists and influencers wandering around London showing off internet speed test results in the 400-500Mbps range.
This is great. It’s four or five times quicker than the fastest home internet available to much of Greater London, let alone rural areas. But this is not the time for most people to invest in a 5G phone. Give it a year or two. Here's why.
The state of 5G in 2019
Today’s 5G phones offer a snapshot of today’s 5G progress. We are only at the beginning of its development and this plays out in a few different ways.
Spectrum is one of the easiest to explain. Phone networks license bands of the frequency spectrum, over which they transmit mobile internet services. These are auctioned off for vast sums of money: EE paid £302,592,000 for 40MHz of the 3.4GHz band. At launch, 5G services are using this 3.4GHz band.
They were auctioned off by regulator Ofcom in April 2018, and were previously used by the Ministry of Defence. These frequency bands are remarkably solid and finite for something “floating” invisibly in the air.
Mobile networks now have some more frequency bandwidth to start rolling out 5G. But even Telefonica UK CEO Mark Evans described this as merely “laying the foundations for 5G in Britain”.
5G will get more interesting when further parts of the spectrum are auctioned off. The marketing tells us that 5G is a tech masterpiece that will solve all the bandwidth issues of the congested 4G networks. But you don’t get this simply by using slices of the 3.4GHz band and early testers report seeing 4G-like speeds in many areas where a phone displays a 5G signal, before the public are even using 5G plans.
More antennas are needed
5G antennas also use "massive MIMO” tech, responsible for the claims it will allow for an exponential increase in the number of connected devices. This is a great step forward, but the decreased wavelength of 5G’s bands also means decreased range for each transmitter. More 5G masts, or smaller antenna clusters, are needed, everywhere. And until we get them, 5G signal will be restricted to fairly small pockets in cities around the UK.
Building the required infrastructure will take years. And it is likely to be slowed both by protests of those who believe the baseless claims that 5G is gradually sautéing their brains, and the current hot water Huawei is in.
Huawei is one of the primary manufacturers of networking hardware, and was a giant in this area before anyone had heard of its smartphones. Views and allegations around Huawei include that it is a shameless IP thief, an unacceptably successful rival to non-Chinese companies, and effectively the Chinese government in a friendly Bill Clinton face mask. (Huawei has denied claims against it).
Many are now incensed by the idea Huawei has a part in the UK’s 5G setup. Bad news for those people: it was already there in our 3G and 4G infrastructure, although BT and EE have reportedly started removing Huawei’s kit.
The signal isn't great yet
Outside a few hotspots, you’re likely to see 4G-like speeds in most places, even in central London, for some time. Even if your phone says you’re connected to 5G.
There’s another issue. The boldest claims about 5G speed rely on the 'mmWave' bands that phone networks don’t have access to, yet. This refers (primarily) to bands in the 28GHz range, frequencies so high that the signal’s wavelength is reduced to a millimetre, instead of the around 90mm of the current 3.4GHz 5G signal.
MmWave is what will provide the real mind-blowing speed of 5G once lots of actual people start using it. There are, you guessed it, more problems. MmWave signals have trouble passing through walls, and the maximum range of antennas, even with line of sight, is around 500m according to Verizon tests in the US. The network is actually using mmWave tech.
If Ofcom, as is expected, grants access to the 26GHz band, we’ll start seeing 5G antennas everywhere. Bus stops, the sides of buildings, underneath the shopping trolleys of unsuspecting old ladies. That said, they’ll likely become immediately invisible like the routers strapped to the ceilings of just about all London tube stations.
5G is important, but it is not simply a switch that is flicked at Vodafone HQ. It’s a process, and a long one.
The phone hardware could be better
Today’s hardware is not quite ready for all the changes either. All 5G launch phones – the OnePlus 7 Pro 5G, Galaxy S10 5G, Oppo Reno 5G and LG V50 5G - have the Qualcomm Snapdragon X50 modem.
This lacks a few features that may be important for the future of UK 5G (and already are in other countries including the US). The Snapdragon X50 does not support SA, required for full standalone 5G networks not paired with 4G, or FDD. This is where upload and download data use distinct frequency channels. The alternative, TDD (used by the Snapdragon X50 and launch 5G services), makes both travel in the same band. TDD makes sense now, when the network 'space' for 5G is limited. But networks started using FDD fairly soon after 4G’s introduction.
Perhaps this doesn’t matter if Brexit slows infrastructure improvements and leaves us huddled around a burnt-out Morrisons, holding up our phones to a 5G mast made of old spoons and a coat hanger. But the Snapdragon X50 is very much a first-wave 5G modem.
Qualcomm announced the X50 way back in 2016, and the upcoming Snapdragon X55 fills all its major holes, and is likely to built into future SoCs as standard. The Snapdragon X50 is not – it has to be added by the manufacturer.
Data is too expensive
The most impactful change we need to see as mobile phone users has nothing to do with infrastructure, 5G hardware or phones. Data allowances are the biggest issue.
EE’s early 5G contracts come with no more data than a high-end 4G one. You can pay £59 a month for a contract with just 10GB of data allowance, which reduces 5G’s appeal to faster-loading web pages. Its £79, 120GB plan is much closer to what we want to see for 5G, but the cost is prohibitively high for most.
To return to the Samsung example: you can get an EE plan for the regular Galaxy S10 (128GB) from £45 a month; the more comparable Galaxy S10+ (128GB) from £50 a month and the Galaxy S10 5G (256GB) from £69 a month with 30GB of data.
If 5G has a 'user experience' purpose, it's to remove the barriers between local and streamed content. So you should be able to stream a movie as if it were a file on your phone. A game played through Google Stadia should seem the same as one installed on internal memory.
Even if the low latency and high speed of 5G are capable of this trick, it’s not much use if you have to constantly worry about how much data you use.
Thanks to the smart home, any vision of having one internet connection for home and away are far off. How will you control your smart thermostat when you accidentally leave it on, and your phone is five miles away from it, in your pocket? In future we might see plans with home 'access points' that share the same allowance as your phone. But any claims 5G will spark a behavioural or major experiential change seem thin when too many of us have to treat our mobile internet like a pre-pay boiler.
5G is a jigsaw puzzle. The pieces already in place look great. They’re shiny and colourful, but it will be some time before the picture looks complete. Early adopters are right to be excited by what is already on offer. But until it's more mature, the 5G experience may be a little too close to that of stumbling across the occasional bafflingly fast Wi-Fi hotspot. 2019 may be the official year of 5G, but in 2020 and 2021 it’s more likely to be ready for mass adoption.
This article was originally published by WIRED UK
