5G. You're aware of it, probably. You've at least likely whizzed past it unblinkingly in a press release or heard and ignored it in a sound byte from a CEO or analyst, as 5G is increasingly a very hot topic in the biz.
But wait, aren't we just kind of now starting to get 4G everywhere? Isn't it too early to be talking about the next generation of wireless technology? Actually: yes. It kind of is. And that's part of what this piece is about - providing a bit of a reality check on 5G, what it is, what it means for you (or doesn't), and when you can expect it.
First: some terms
I just want to lay out some definitions to keep things clear here. These aren't dictionary, technical, or Wikipedia definitions of these terms, but they're more about the way they're used in this piece.
- 3G: Wireless networks of the third-generation, which you may also know by their most-recent data access revisions, HSPA or HSPA+.
- 4G: For the purpose of this article, 4G refers to LTE-based networks and LTE-based networks only. The terms here are completely interchangeable. LTE, LTE-Advanced, LTE-Advanced Pro, this is all "4G."
- 5G: A term used to describe the "next generation" of wireless communication networks that have yet to be standardized.
- Millimeter wave: A term used to describe portions of the Super or Ultra-high-frequency spectrum, particularly those frequencies above 6GHz (so, starting 1GHz above where your 5GHz Wi-Fi router operates). 5G networks are widely anticipated to use millimeter wave frequencies, but for the purpose of this post, when you think 5G, think 6GHz+, and when you think 6GHz+, think millimeter wave.
- Mbps/Gbps: Megabits and gigabits per second, respectively. 4G was originally defined as requiring end user connection speeds of 100Mbps (less known: 4G "fixed" access speeds were supposed to be 1Gbps). 5G has been defined by some as 1Gbps for mobile devices. These definitions are, in all likelihood, pretty meaningless for end users, and just serve as goal posts for mobile operators and network equipment vendors.
- Carrier/mobile operator/wireless provider: Any company in the business of selling access to a wireless network - such as Vodafone, Verizon, T-Mobile, etc. One day, this could also include your cable, satellite, or landline ISP company - though these companies have long made big talk of getting in on wireless, and have rarely actually done so.
- Spectrum: The electromagnetic frequency spectrum, measured in Hertz, which includes things like radio waves, microwaves, VHF, UHF, etcetera. Current cell networks operate almost exclusively in the ultra high frequency spectrum, which extends from 300 to 3000 Megahertz. 5G networks are anticipated to operate in much higher frequencies, in the super and extremely high frequency parts of the spectrum, especially those frequencies between 6 and 100 Gigahertz.
- Latency: the time it takes information to travel between points, often in terms of "round trip" times - the amount of time between the moment a "ping" is sent from an end device to a point on a network and the moment a response "ping" is received from that point. In networks, this is typically measured in milliseconds. 5G networks, supposedly, will offer even better device-to-network (smartphone to "tower" or whatever the network access point is) latency, possibly even under 1 millisecond for fixed communication.
- Internet of Things: An annoying but useful term to describe all the various gadgets and devices - everything from smartphones to refrigerators to fitness bands to cars - that are increasingly connecting up to and interacting on the internet and cloud. There is expected to be a huge increase in the number of such products in the future, and 5G is being envisioned in part to respond to this massive influx of new connected devices. Abbreviated as "IoT."
What the heck actually is "5G"?
That's an excellent question - and unfortunately there is no decided answer to it at this point. I know that's frustrating, but 5G is very much in a fluid state, and no one is quite sure exactly what it's going to be yet. The best way to describe what 5G is is to describe the problems it hopes to solve, and I know that's not the easy answer you want, but it's really important if you do want to get a grip on the concept. Obnoxious buzzword-ridden videos like the one below from Intel or this one from Samsung don't help the situation.
This pointless 5G video from Intel will teach you exactly nothing.
Right now, wireless networks, especially in developed countries, face a real problem: congestion (granted, 4G was supposed to fix that). I live in one of the world's largest metropolitan zones (Los Angeles), and I can tell you: 4G LTE in LA sucks. T-Mobile and Sprint actually provide decent LTE service here [outdoors, in some areas], but that's basically a result of their not being as congested as AT&T and Verizon yet and having moderately newer networks, not them having some LTE silver bullet that other carriers don't. My AT&T LTE speeds at peak hours are sub-1Mbps up and down - I literally switch to the less-congested HSPA+ (3G) network sometimes because it performs better.
This isn't a problem specific to my region, my state, or my country: it's happening all over the world. And we do have contingency plans in place to help existing 4G-LTE networks scale with all this increased demand - for example, check out 4.5G (known as LTE-Advanced Pro, yes really, that's the name).
But mobile operators believe data use and the number of network-connected devices (the "Internet of Things" is a common umbrella term for such products) are going to increase exponentially in the next decade. And there is a strong desire to move quickly to implement technologies that will allow wireless networks to scale to meet the demands of that growth. You know, before they get congested into oblivion by all of our damn cat videos streaming in 4K while we're simultaneously uploading a bunch of 20MP photos to the cloud and buffering the next track in our lossless music library while our 38 connected cameras and locks and thermostats and light bulbs sit idle in our home eighty miles away doing internet of things stuff and sending data back to our phones. Basically, we're all data hogs, and we're increasingly hungry. And there are more and more people living most of their "digital lives" (related: UGH I HATE THAT TERM) on their smartphones while utilizing a mobile network.
Future improvements to 4G network technology will help address this in some ways that, frankly, are very complicated and I really don't quite understand and will not pretend to for the purpose of this article. But the short of it is this: current wireless networks and, more specifically, the frequency and bandwidth they're limited to, weren't built anticipating the truly massive amount of data and connection demand carriers are starting to see and are now extrapolating for in the years to come.
5G is, hopefully, the light at the end of the tunnel for scalability: the ability to have a network that can handle truly massive amounts of data throughput in highly congested areas or during peak usage hours. The reason it will be so much better at this? Most carriers and network equipment research has placed the bulk of 5G's initial promise in operation at frequencies exceeding 3-6GHz, often well above that - 28GHz is a popular choice right now. These frequencies allow for tremendously increased data capacity and speed for reasons I do not fully comprehend, but a big part of it is that "spectral efficiency" increases greatly at super/extremely-high frequencies. Current mobile networks, by comparison, generally operate in the 0.5-3GHz range. The downside to these much higher "millimeter wave" frequencies is that they have far poorer range than current cell networks.
And that's why 5G also isn't all about increasing your end connection speed - a lot of 5G is concerned with making the network faster and more efficient from a fixed standpoint (towers talking to towers, towers talking to buildings, etc.). In fact, it's very possible networks will be communicating with 5G long before your smartphone does. So, let's move to a hypothetical to try and explain what 5G might look like in the real world, because it's not quite as simple as "your phone now shows 5G in that status bar and downloads stuff stupid-dumb-wicked-5G-fast."
What 5G could look like
You're in a building with 1000 other people, and the work Wi-Fi blocks anything that could remotely be considered "fun." Everyone is using the same wireless carrier during lunch to watch Netflix. You're all on an LTE signal that's coming from a tower a half-mile away, even though you're all in the same building and could probably be utilizing a much more efficient short-range, high-speed connection. You're also congesting the mobile network for everyone who needs mobile access to that tower, and that slows things down. If you know a bit about networks, you may already be thinking "small cell site" - a device that connects to a hardline in the building out to the street for internet access and that emits a wireless signal your cell phone can use. Sort of like a wireless access point, but for cell signal.
5G could, in theory, take this concept even further, and eliminate the wireline network requirement of small cell sites and greatly increase speeds and capacity in the process. A millimeter wave 5G repeater could be mounted on top of this building (or a nearby one). Via the building's existing network infrastructure, access to end devices could be provided via a simpler 5G "picocell" (on a separate virtual network) inside the building or even existing Wi-Fi APs, which would then in turn route traffic to the 5G repeater via those existing wired connections. Traffic routed to the 5G repeater on the roof would then be transmitted [insanely quickly] to a nearby 5G cell tower (or another nearby bridge that could reach one) with full fiber line access and all the necessary networking equipment (likely at the site of an existing 4G tower). This would greatly reduce the congestion on the valuable low-frequency 4G bands, which are necessary for outdoor to indoor building penetration and covering large swathes of regions.
AT&T's Project AirGig is one very interesting vision for a potential way 5G could be initially deployed.
In theory, 5G could see the adoption of a much more distributed network model in extremely dense urban environments. Instead of having to lay fiber everywhere or tap into existing but bottlenecked landline networks, 5G could act as a wireless network bridge of sorts. And because of the tremendous bandwidth and speed 5G's super/extremely-high frequencies allow (along with extremely low latency), a cluster of millimeter wave 5G access points could theoretically provide connectivity to a huge number of devices in a very small area.
5G could just as easily be an always-available outdoor network in high-traffic areas in major cities - think Times Square - or along large freeways (cars increasingly utilize mobile data). The kinds of areas where we can relatively precisely know where end devices will be concentrated, and know that there will be a very large quantity of them.
There are many ways the advantages of 5G's extreme speed and capacity could be leveraged even given millimeter wave's limitations on signal penetration and propagation (in essence, how far the signal goes). There's a good paper here on the nature of millimeter wave's signal interaction with materials, and it turns out while it's not well suited for penetrating into a building from an outdoor antenna (tinted glass and metal are highly reflective of millimeter wave signals), 5G does have good "indoor-to-indoor" penetration, which could allow it to serve as the backbone for mobile networks in buildings or other enclosed areas. This is important because we are currently experiencing what many carriers are calling a "spectrum crunch." There are only so many parts of the EM spectrum from around 500MHz to 3000MHz available, and we're already using a lot of them. Our current networks are constrained by this lack of spectrum (in theory), and 5G's benefits could dramatically reduce the strain on networks in these frequencies where we need it most - in high-congestion areas. 5G could also have applications outside high congestion locations, but that's just one example.
You could also have a 5G access in your home as an alternative to a traditional ISP, with a 5G receiver mounted on your house - Verizon will be trialing such a thing. And another thing to remember about all these hypotheticals? Wireless carriers are very much interested in ways they could potentially charge you more for 5G or steal business from traditional wireline ISPs, and that probably will lead them to posit hypothetical business or network models that actually end up making no practical sense whatsoever. We won't know how carriers actually will use 5G until they announce hard plans and begin rolling it out in the real world. Right now, 5G's actual end user experience is just a lot of talk: we don't know what it's going to look like, we can just theorize ways the benefits of 5G networks (at least right now) might solve problems.
Which leads us to the other, inevitable question you no doubt have.
Yeah, but how fast would 5G be really?
Listen, we already got sold the kool-aid on 4G being "defined" as 100Mbps on mobile devices. That should already have you cautious of promises that 5G will provide 1Gbps to end users. That seems... ambitious, at least for now.
More likely, 5G - when and if it is widely-deployed - will first focus heavily on simply making wireless connectivity suck less. A congested LTE network is almost no better than a congested HSPA+ network. When both are under such heavy loads that speeds come to a crawl, they both provide a bad user experience. 5G's "fat pipes" in this new, highly efficient part of the EM spectrum will hopefully reduce the likelihood of crippling congestion happening in the first place, because it's being envisioned from the ground up for these extremely high-demand, high-volume applications.
And yes, 5G is meant to support ridiculous, fully ludicrous wireless data transfer speeds - Ericsson has achieved field test results of 25 GIGABITS per second. How fast is that? Your laptop or desktop PC may have a SATA 3 interface for storage. That interface peaks at speeds of 6Gbps, and the fastest consumer solid-state SATA drives at the moment top out well under that (though, admittedly, new cutting-edge NVMe SSDs are much faster). Anyway, 25 gigabits per second is stupidly fast.
Samsung offered a demo of its own 5G tech operating at 1.2Gbps - while traveling 100km/h.
And end device speeds would be nowhere near this. Such extreme speeds are mostly technical and marketing demonstrations, and at best serve to illustrate how fast wireless network-to-network communication could be in a fixed, highly-controlled application. So every time you see a new article about some company breaking the 5G speed record or talking about speeds greatly exceeding 1Gbps, you should mentally filter that out as "probably not applicable to me." It may well be exciting technical news in the industry, but these things really have no direct bearing on you, the consumer and user. They may have indirect benefits (i.e., even greater network capacity), but that's another discussion.
The real answer is we don't know how fast 5G will be. Transmission tests and theoretical maximums exist in a vacuum - the real world is a much more complicated and messy place full of equipment bottlenecks, practical concerns, and cost considerations. 5G's insane speeds are, for now, much more about building wireless networks that will scale well when those various real-world bottlenecks start to become more practical and less expensive to widen.
So we're going to replace 4G with 5G?
If you've read this far, you probably can guess the answer to this question: yes and no. And we don't know. Yet. (I suspect you're noticing quite a theme in this article, by the way.)
5G isn't explicitly meant to replace 4G immediately. And it's actually not clear if it will replace 4G at all in practice for a long time yet, or if 4G-LTE technology will be absorbed into 5G in some way, or more likely, if it's a combination of the two in some way that will involve a very exhaustive set of meetings between people who come up with obnoxious technical marketing buzzwords (4G+5G=9G? BRILLIANT!). People making lots more money and with many more degrees than me are all still figuring this stuff out from the technical end, and tongue in cheek comments aside, I cannot claim to understand the many intricacies they must be considering.
Qualcomm believes early 5G and "4.5G" will coexist for some period of time.
What I can say is that because of the poor signal range and non-existent building penetration of the 6GHz+ signals on which 5G is currently focusing, it is very likely 4G and evolutions of 4G will be around for a very, very long time yet (like, at least a decade). 5G, as currently imagined in these super high frequencies by carriers like Verizon and AT&T, is basically the antithesis of the sprawling coverage maps we've created with cell networks to date. These 6GHz+ networks would be highly focused, require lots of smaller localized access points (as opposed to a few huge towers), and would probably have very sharp signal boundaries. The moment you leave a building that has millimeter wave 5G access points, you'll probably lose the 5G signal, unless outdoor coverage is available. This is in sharp contrast to the situation with 3G and 4G, which seek to "blanket" a large area with signal often using a few strategically placed towers.
But in very dense urban areas, 5G may well supplant 4G fully in effect, with 4G acting as a fallback for indoor coverage where 5G access points aren't available and fringe locations that 5G signal doesn't reach. Again, you can probably see the point here: taking the load off these valuable frequency bands that do actually penetrate buildings and which provide that "always-on" coverage we all need sometimes. Much like the early days of 4G. Though it seems likely that 4G and 5G will see an effective coexistence much longer than 3G and 4G did, simply given how ambitious and different 5G is. Qualcomm doesn't anticipate 5G to really come into play until "2020+" on its technology roadmap.
At the earliest, Qualcomm pegs 5G for a late 2019 launch.
And by then, LTE networks will be very mature - substantially more so than they are today. Believe it or not, LTE still has a lot of technical upgrades that are nowhere near implementation out in the wild yet. LTE-Advanced Pro, sometimes called "4.5G," is the next major evolution of LTE, and actually incorporates some 5G concepts such as use of much higher signal frequencies and dubious promises about gigabit speeds. So, the lines between 4G and 5G are already kind-of-sort-of getting blurry, and this new "4.5G" generation of networks isn't even expected to see major rollouts for another few years. 5G, then, at least for consumer access, is still way off on the horizon.
Who is going to have the first "5G" network, and when?
Great question! It's anybody's guess. Even with the 3GPP working to define 5G, there's nothing that stops some random carrier from claiming tomorrow that they have a "5G" network. You know how it goes with marketing: the guy who claims to have the amazing new technology first is probably stretching the truth. The 3GPP only finalized "4.5G" back in March of 2016. They don't expect 5G's specification to be fully set in stone until 2020. Granted, the first release should happen in 2018, but possibly slightly sooner if carriers can prod the 3GPP into moving more quickly.
Again, that doesn't stop anyone claiming they have a "5G" network in the meantime. 5G isn't a protected term, and as such, will be subject to the abuse of companies who want to be seen as being on the cutting edge of mobile technology. But as consumers, we probably shouldn't expect "true" 5G networks as defined by the 3GPP until the early 2020s. And that's a long way off.
So, that probably dampens your 5G enthusiasm a bit, and I'm sorry about that. With all the hubbub about 5G testing and speed records being broken, you'd think we were on the cusp of seeing these things ready to go and ramping up for inclusion in the next few generations of smartphones. Unfortunately, that's unlikely. But on the bright side, 4G LTE networks still have lots of growing to do, and the changes brought about by the Advanced Pro (4.5G) specification are not insubstantial. In fact, they're massive - because 4.5G is all about laying the foundation for 5G. Here in the US, carriers don't even offer the full suite of LTE Advanced features yet, so American consumers haven't even seen what Advanced Pro can offer them (granted, no one else has, either). LTE Advanced Pro aka 4.5G really is the next big evolution of mobile networks we're going to see, but I wouldn't be surprised if, like many late-stage 3G technologies, it gets "up-branded" to 5G.
In the meantime, just remember any time you read about 5G that what you're seeing is one company's interpretation of 5G. 5G will likely evolve and change considerably by the time the specification is finalized in 2020, and in that time we'll almost definitely see more practical and realistic imaginings of what a 5G network will look like for end consumers. Until then, expect to see a "4G" (or an LTE-A/Pro) in your status bar for years to come.