What is 5G?

The 2017 Mobile World Congress trade show kicks off next week, and in addition to the plethora of new smartphones, 5G network news is expected to show up in a big way. But what exactly is 5G? Is that the same as gigabit networks? LTE Advanced? Is the whole thing just a marketing trick, like when AT&T and T-Mobile renamed HSPA+ as “4G” data to cover for their lack of LTE support?

What is 5G?

In the simplest possible definition, 5G is the fifth generation of cellular networking. It’s the next step in mobile technology, what the phones and tablets of the future will use for data, and it should make our current LTE networks feel as slow and irrelevant as 3G data seems now.

To recap, the first generation of mobile networks (retroactively referred to as 1G) came out in around 1982. It was a fully analog system until the launch of 2G (second generation networks), which made the jump to digital when it launched in 1991. 2G also added cellular data in the forms of GPRS and EDGE technologies. Roughly 10 years later, 3G networks launched, offering an even faster data rate than 2G. Around 10 years after that, our current LTE networks — what we call 4G, although there’s some contention on what that really means — is the fourth generation of networking. Historically, that works out to a new generation of networking technology every decade or so. 5G networks will presumably offer a similar leap forward when it comes to things like data speed.

When is 5G coming?

Working off that model, in the best case scenario, we could see commercial 5G phones in the early 2020s, assuming the same “every 10 years” pattern as previous generations holds through. LTE began to roll out (at least in the United States) in around 2010–2011, so some simple math shows that we should expect to see 5G in 2021, which is coming up quick. Chances are, we’ll see some earlier deployments even sooner than that, if the network providers, modem manufacturers, and wireless carriers are able to live up to their early projected roadmaps. Qualcomm plans to make its early 5G products available to the public as soon as the 2018 Winter Olympics in South Korea. Like the jump from 3G to LTE, you will need a compatible phone to take advantage of 5G when it does roll around, but you’ve still got a few years to figure that out — obviously, we’re not expecting to see any 5G phones launching at MWC this year.

How is 5G different from 4G?

The most important thing to know about 5G is that there is no official “5G” yet. No matter what we hear at MWC this year, no matter how fast the speed test demos, or how different the networking technologies that companies use are, 5G is still a glimmer of an idea in the distance.

A 5G network will have specifications beyond those for 4G, but it hasn’t even been been agreed upon yet what those technical goalposts should even be (there is, however, a logo.) As former FCC chairman Tom Wheeler noted last summer, “If anyone tells you they know the details of what 5G will deliver, walk the other way." Expectations for commercial 5G range from internet speeds in the gigabit or even tens of gigabits range and vague goals of lower latency, but at this point in time we simply don’t know what 5G will truly look like.

That said, there are some ideas of what we can expect. Companies like Verizon, AT&T, Intel, and Qualcomm are already spinning up tests for 5G technology, and it’s these early experiments that will likely shape what the formal international standard for 5G becomes. One of the commonly cited features for 5G is the use of millimeter wave (mmWave) band transmission, which could be the key to unlocking the blazing-fast internet speeds that 5G promises.

What is mmWave technology? Why is it better?

Cellular technology transmits data over radio waves, which depending on the type of electromagnetic signal is measured as a different frequency. The higher the frequency, the smaller the wavelength, so millimeter wave technology refers to signals with a wavelength that’s measured in millimeters, and is generally defined as between 30 GHz and 300 GHz. For 5G, the FCC has already made available swaths of the spectrum in the millimeter wave range for both licensed and unlicensed use as of last summer for companies to begin exploring 5G options (specifically, licensed use in the 28 GHz, 37 GHz, and 39 GHz bands, unlicensed use in the 64-71 GHz band, and shared access in the 37-37.6 GHz band).

Why do we care? Because millimeter wave technology promises higher data capacity than we currently have now. A simplified rule of thumb to go by is the higher the frequency, the more data it can transmit. So, FM radio, which transmits just audio, typically broadcasts at between 87.5 to 108.0 MHz, but LTE — which is responsible for far larger data — streams between 700 MHz to 2,100 MHz (i.e., 2.1 GHz). Millimeter wave technology would offer the bandwidth for orders of magnitude of improvement over LTE. We’ve already even seen commercial use of millimeter wave technology in things like the Starry Beam. (This trend continues up the electromagnetic spectrum into visible light, which has a frequency between 430–770 THz — that’s up to 770,000 GHz — which is one of the reasons why fiber optic technology is so fast.)

Another advantage to the shorter wavelengths found in millimeter wave technology is that antennas used to transmit and receive the signals can be made comparably smaller. That means that phones that use millimeter wave technology could take advantage of multiple antennas for different millimeter wave bands in a single device, which could result in a more efficient use of the available spectrum and faster internet when multiple users are connected.

Millimeter wave technology comes with its own challenges, however. With higher frequencies comes shorter transmission ranges, and shorter wavelengths tend to experience greater issues when there’s no direct line of sight, along with interference from walls, buildings, window panes, and even raindrops. Whereas older radio and cellular technology were able to rely on a comparatively smaller amount of larger antenna towers, millimeter wave would need lots of smaller antennas peppered around cities and countries to function well. It’s technological issues like these that the early 5G tests will be looking to explore and solve.

Gigabit LTE / LTE Advanced / LTE Advanced Pro (or, I want it now!)

Usable 5G technology is still years away, though (again, there isn’t even a defined specification yet). And while LTE doesn’t deliver gigabit speeds, it’s possible that LTE Advanced and the recently finalized LTE Advanced Pro might serve as a stopgap. LTE Advanced is already available on a variety of phones, and carriers like Verizon, AT&T, and Sprint are beginning to support it on their networks. LTE Advanced Pro is the next evolution of LTE that might make practical gigabit mobile internet a reality, as well as begin to lay the groundwork for technologies that for 5G, including things like MIMO (multiple antennas) technology and use of unused spectrum in the 5 GHz LTE-U band. LTE Advanced Pro is also being set up to be a more widespread alternative build on existing technology to offer potentially gigabit level speeds for when 5G rolls out, to ensure a similar networking experience when outside of the fledging 5G areas (similar to how HSPA+ 3G networks helped bolster connectivity while LTE was rolling out).

LTE Advanced is also taking advantage of a technology called carrier aggregation as a stopgap for existing LTE to reach higher speeds. It works by allowing a device to use multiple LTE bands simultaneously to allow for increased bandwidth, and therefore, increased speed. LTE is theoretically capable of aggregating up to five channels for the best speed rates, but the most we’ve seen on the market yet is three-channel aggregation, which Sprint recently rolled out.

The 4G problem, or tempering expectations

It’s also worth remembering to temper expectations. While on paper, LTE Advanced could offer gigabit speeds, and LTE Advanced Pro is specced for up to 3 gigabits per second, that almost certainly won’t translate directly to the real world. In fact, what you (and cellular network marketing departments) think of as “4G” or LTE isn’t really 4G according to the agreed upon standards from the International Telecommunication Union (the ITU) and the 3GPP. Per those standards, a 4G network (among other things) would provide a 100 megabit/s data rate when moving and a 1 gigabit/s while stationary, something that our current networks certainly aren’t capable of yet (LTE Advanced and Advanced Pro are hoping to succeed as the true candidates for a “real” fourth-generation mobile network.) So it’s possible that the dreams of gigabit LTE and 5G may not quite pan out as promised, or if they do, that the timetable could be longer than expected.

Where do we stand now?

Well, going into MWC, expect to hear a lot of news about ongoing 5G developments as both network and hardware companies work to have the technology in place (whether it turns out to be millimeter wave or something else entirely) to build a true fifth-generation network. We’re already starting to see news in that vein this week — Verizon announced plans for 5G testing with millimeter wave hardware in multiple cities across the US, AT&T is planning to test its own more unspecified “5G Evolution” network later this year, and Intel, Qualcomm, and Samsung all announced new chipsets that can support gigabit LTE speeds.

So while the exact details of future cellular networks — whether LTE, 5G, or beyond — may still be a little hazy, there’s one thing we can say for certain: the future will be fast.

Recommended by Outbrain

Comments

Thank you for publishing a well-written and informative technology article void of politics.

Logged in just to rec this. Especially the "void of politics" part.

Great article.

I’d much rather be able to use my phone inside my office building (at any speed) than get gigabit speeds within line of sight of a tower.

Yes. It seems impossible to me for these ~30GHz (Ka band) to have any real application for mobile networks with the capabilities we have now.

A cynical person would say the only reason thee high frequencies are used in lab settings is to claim large throughput numbers.

In fact going to higher frequencies is the exact opposite of the planned strategy of freeing up the analogue TV bands to built data networks in lower frequencies with better range and penetration properties.

With higher frequencies comes shorter transmission ranges

Minor quibble. Frequency has no impact on transmission range. In fact, in RF there is no transmission range. It’s usable signal depending on several factors. Background noise floor, and the medium (atmosphere) can cause more attenuation of a higher frequencies usable signal at the receiver level. Of course also receiver sensitivity, antenna geometry (at both the transmission and reception side), and transmit power are the other factors that affect the usable signal.

Thanks for drawing out this point. The quibble I have is more specific than your take. The term "transmission ranges" in this context actually mean "signal penetration." That’s how I read that phrase in the article.

But the problem is this is just not true. Signal penetration doesn’t get worse as higher frequencies, it totally depends on the material doing the absorption. For example gamma radiation (like 300 Exahertz) penetrates just about everything not made of lead or couple feets-thick of concrete (obviously this is ionizing radiation so we can’t use it for consumer wireless transmission but that’s besides the point). Ground-penetrating radar uses 30MHz-30GHz range, which is not a coincidence that it’s similar to the frequency of what we use for wireless transmission, as most buildings and hills/mountains/etc is made of similar molecular material. It would be misleading to imply if we go lower frequency-wise, somehow signal transmits better.

Lovely writeup – I miss this Verge, more of these articles please!

5G is the next LED TV: Bullshit marketing term waiting at the gates to be unleashed on the masses who don’t even know they don’t have true 4G yet. Thank you for this article for explaining the technical basis why we break cellular standards into evolutions (since test, rollout, adoption etc take a while).

Indeed. We’re on EE here in the UK, and whilst when you have 4G, the speed is very fast, when it switches to 3G or 2G, the data rates are unusable because they knobble 3G and 3.5G to get 4G. (Deep technical reasons behind this.) And the strong signals are in the wrong places, such as city centres, where you don’t need it because you will either be inside using WiFi, but when outside walking around, all you need to do is talk or stream music. But if you’re on a train or in a car, you will want high speed wireless so you can work online, such as with Google docs and such.

Service with Verizon has actually gotten worse in places like Brooklyn and semi-rural Pennsylvania since the switch to LTE.

Their service maps are misleading considering many areas listed as "covered" have a few bars when you’re outside, but the moment you step foot indoors, you’re lucky to be able to send a text message.

I know they’re not listening (because since when have they cared about customers?), but it would be nice if they could just have truly good coverage before investing in an another new standard.

If I can clock a steady 100Mbit I’m more than happy. I can’t think of any mobile application that requires more than that today.

The logo looks like it is from the 90s.

As you noted in a couple places, "4G" really is more of a marketing term than anything else. According to the 3GPP (which is itself pretty confusingly named – it’s a group that came together to try to make all then-futuristic 3G networks work together – it’s short for 3G Partnership Plan), "4G LTE" isn’t usually 4G by speed, but 4G is also defined by other characteristics (like Orthogonal Frequency Division Multiplexing (OFDM) to increase speeds), which LTE does have.

The article starts out with HSPA+ being mentioned as "faux 4G," but in reality, T-Mobile’s implementation was considerably faster than Sprint’s "actual 4G" WiMax network, and at least in the same ballpark as AT&T/Verizon’s LTE in terms of actual download speeds. (Plus, Verizon needed LTE more – since it’s 3G network couldn’t handle simultaneous voice and data, which T-Mobile and AT&T’s GSM-based networks could.)

The real benefits of 5G – which obviously haven’t been defined yet – aren’t likely to be so much about the speeds of an individual phone, but rather, an increase in the number of devices that can use data at high speeds in a small area – in other words, less congested networks.

The real game changer is latency above all else. Will allow cloud computing to make any device infinitely more powerful.

According to Ars, the ITU has published a draft spec for 5G meaning it’s a little more fleshed out than this article claims.

https://arstechnica.co.uk/information-technology/2017/02/5g-imt-2020-specs/

A few things to point out here just for clarification:

  • mmWave is not "generally" defined as 30 GHz to 300 GHz. The actual wavelength of the signal is from 10 millimeters to 1 millimeter (wavelength = speed of light/frequency). Yes 28 GHz gets lumped into mmWave bands, but there is no "general" when it comes the math of determining the wavelength
  • What your forgot when mentioning 64-71 GHz is that 57-64 GHz is already unlicensed! That’s a total of 14 GHz of contiguous bandwidth for unlicensed use. Although, 60 GHz is in the oxygen absorption band which causes severe attenuation.
  • It’s not just the higher in frequency you go the more data you can transmit. It’s that there is much more bandwidth around carrier frequencies compared to low microwave. Think of it this way. All of the systems operating at 5 GHz or below could easily fit into the 14 GHz of bandwidth centered at 64 GHz. Not possible due to technology limitations and propagation implications at the 64 GHz band, but just think about that!
  • The Starry router has not been proven yet besides rare publicity occasions with reporters given some access. They are operating at 38 GHz and have NOT proven their claims for multi-gigabit to multiple access points from one base station, even in NLOS. I would like to see more proof-of-concept or numbers to give credence to their system and early claims.
  • Buildings, walls, and window panes do NOT cause interference, they can actually work as great reflectors for mmWave. What you mean is that mmWave does not penetrate well through materials compared to current cellular bands. This could potentially be a good thing though, where systems can take advantage of interference isolation.
  • Rain over kilometers will cause issues, but not in a small cell sense (200 m cell radius), which would experience a few dB of additional attenuation for moderate to heavy rainfall.
  • Ericsson is looking above 100 GHz now for backhaul and wireless solutions, even before 5G below 100 GHz rolls out: https://www.ericsson.com/assets/local/publications/ericsson-technology-review/docs/2017/etr-beyond-100ghz.pdf
  • More antennas can fit in a smaller form factor at mmWave, this is true, but this is also needed in order to overcome the additional attenuation and free space path loss experienced at higher frequencies that current cellular bands do not experience. For instance in the first meter of free space propagation, the 700 MHz band experiences 30 dB of attenuation. The 28 GHz band experiences 61 dB of attenuation in the first meter of propagation. That means right off the bat (at 1 m), the 28 GHz signal is 1000 times weaker than the 700 MHz signal (assuming each has an antenna size on the order of the wavelength). Increasing antennas at mmWave is possible, but this makes the antenna beams much more narrower in order to have gain and directivity to maintain links with reasonable SNR. A downside to this is the effect of blockage by trees, cars, people, and other objects in the environment. Current cellular mobiles operate in an omnidirectional sense where the antenna can receive energy from all angles at one time, whereas at mmWave, the antennas will have to actively steer in the direction the signal is coming from. This will require efficient DSP algorithms and will come at cost, but will eventually happen.
View All Comments
Back to top ↑