One of the events I attended at CES was a panel discussion on 5G, the umbrella term for the next generation of cell phone network technologies. Industry watchers don’t expect 5G to be widely available until 2020, but work is already underway to develop the necessary technology and standards.
Historically, new cell technologies are faster, and prices will be high at first but will eventually come down to “normal,” with different people deciding what’s reasonable to spend. 5G will likely be different, however. Yes, it will be faster, breathtakingly faster, in fact, with peak speeds of perhaps 26 gigabytes per second. (Yes, that’s bytes, not bits.) But 5G will be better in numerous other ways too, which could make what we’re using now seem rapidly antiquated, in much the same way that a modern broadband connection isn’t merely faster than dial-up.
This raises a chicken-and-egg question, though: what exactly would we do with networks that are this fast? Do we really need streaming 4K video on our phones? Will we look back on our current video technology the same way one might regard a first-generation 3GP video today? What other media and Internet services become available when such throughput is widely and reliably available?
I’ll provide an overview of the session and then bring us back to Earth with an analysis of whether we’ll get what’s being promised. The irony of the session is that we were discussing multi-gigabyte cellular throughput in a hotel conference room that promised free Wi-Fi at 384 Kbps.
384 Kbps hasn’t been considered a decent Internet speed for years, and selling 1 Mbps service for $15 per day is the very definition of “shameless.” This disconnect with reality was good cause for skepticism in the room.
Goals for 5G — No one on the panel talked about 26 gigabytes per second; that came from the descriptive materials that CES published before the event. Interestingly, the Wikipedia page for 5G says that faster speeds are not a goal, so I’m unsure where the Consumer Technology Association came up with it. Is it a technical or marketing goal, or something that just naturally pops out of technology in current planning? Why not 8 GBps, or 100 GBps? Pretty much any number in front of “gigabytes per second” is a major improvement on what we have now.
But it’s not just about speed. 5G is designed to have much lower latency than current wireless services. Latency is the amount of time it takes for the network to figure out what to do before it actually happens; if bandwidth is analogous to the amount of water coming out of a garden hose, latency is the amount of time between turning the spigot and the water first spurting out of the hose. Some latency is unavoidable because the speed of light can be annoyingly slow for some purposes, but we humans perceive actions as instantaneous when latency takes only a few milliseconds. This could blur the distinction between what’s in the cloud and what’s on our devices; if we download something the perceived instant we press a button, what
difference does it make if it’s local or on a server?
5G will also support extremely low-throughput, low-frequency data that will be useful for Internet of Things devices. On a 4G network, a network-enabled sensor needs to be on the network at all times, which kills its battery life. 5G will include protocols that use far less power, enabling devices that can run for years without a battery recharge or replacement. In keeping with the belief that far more of what’s currently not “smart” will be network-enabled in some way, 5G networks will be able to handle far more devices on a single cell simultaneously.
Several panelists referred to 5G’s capability to do this as “network slicing.” I’ve been unable to find a reference comprehensible to the layperson on how this magic takes place, but in effect, a network slice is a guaranteed slot of airtime and bandwidth to a 5G device granted by the cell tower. You don’t need to worry about dropped calls or temporary Internet data brownouts; the connection to the network is a guarantee of high service quality. This in turn allows for different expectations of what can be done with the network. For instance, it’s not a big deal on 4G if you can’t get through to Netflix for ten minutes, but
it could be a bigger deal if a hypothetical robotaxi loses connectivity for that time. (That said, increased network reliability makes it easier to sell time-limited services like movie rentals if no one is concerned about download time or jittery connections.)
In theory, the promised nirvana of 5G is that the network is so fast and so reliable that we forget it’s there. If you’re not sure how different that is from today, think about how often you check the bars on your phone for signal (or Wi-Fi bars in your Mac menu bar). 5G hopes to be so reliable that those indicators can go away. Ask yourself how certain you’d need to be about network connectivity before you’d be willing to let those bars disappear.
5G Timeline — For devices and software from hundreds of vendors to interoperate effectively, a standard must first be drafted so that everyone can build on the same platform. These standards are recommended and issued by a cornucopia of acronym agencies, including the Next Generation Mobile Networks Alliance (NGMN), the International Telecommunications Union (ITU), the 3rd Generation Partnership Project (3GPP), and the Institute of Electrical and Electronics Engineers (IEEE). It’s unclear to me which, if any, organization has the final say on the standard. Regardless, standards are voluntary, so it’s more a question of getting everyone to agree on technological
protocols than bureaucratic ones.
The panel was consistent in suggesting that 2020 was going to be when most people (presumably, in technologically advanced nations) would get their first taste of 5G; Wikipedia is less optimistic, suggesting the early 2020s for when the final standard would be released. Draft standards will exist between now and then, but because the drafts can and will change, those buying hardware (or companies investing in networks) may suffer an early adopter penalty if their devices aren’t fully compatible with later drafts or the final standard. So while a 5G modem chip is already available, don’t rush out and buy devices using it.
5G deployment may be delayed in the United States for two additional reasons. The first is network investment; our telecom carriers invested heavily in 4G LTE networks, and they may balk at moving to 5G until they’ve effectively recouped that investment.
Second, as with earlier networks, rollout in the United States might also be delayed by its vast rural geography, as well as relatively light regulation for cellular networks, whereby the major carriers decide their own plans for upgrades and competition. Nothing that I’ve heard of in 5G allows for wide-area coverage from a single cellular tower, so urban areas will almost certainly get 5G coverage first — along with major arterials between cities. Less-populated areas will lag behind.
Beyond that, of course, your devices must support 5G before the fancy network does anything for you. Some manufacturers will introduce it early to compete on technology, but if Apple isn’t among them, you’ll have to wait longer for a 5G-enabled iPhone.
Depending on the cost of 5G chips during the early adoption period, you may see more 5G networks deployed to compete with existing home cable and fiber networks, with standalone modems coming in more cheaply than those small enough to fit in a cell phone. (This would be excellent for people currently living under Internet monopolies, but again, the less dense your population, the less likely you’ll see rapid 5G buildout in your region.)
In any case, South Korea’s SK Telecom has committed to having a test 5G network up and running at the 2018 Pyeongchang Winter Olympics, so you’re about two years away from hearing breathless news articles about a network technology you won’t be able to use yet.
Potential Uses of 5G — It would be wonderful to live with a technology where we truly forget about networks and coverage, and everything just works all the time. When I’m traveling, I ritually check my cellular network signal in every new hotel room and then compare that to the speed of the local Wi-Fi. In most large hotels, the quality of both varies widely based on room assignment.
I can see a near-perfect network being possible in urban areas, but I don’t see how it could work as well in the countryside or tricky locations like subways and elevators. Barring some breakthrough technology for piercing radio dead zones, the laws of physics will have their say. 5G networks will allow for simultaneous use of multiple radio frequencies, and some are better than others for penetrating substantial physical barriers, so you can expect some improvement. But in general, some frequencies are good for lots of data, and others are good for lots of distance or obstructed connections to a cell tower. If you’re in a zone where you’re relying on lower-frequency penetration of lots of physical objects, you won’t get the
Vastly higher maximum speeds might also enable new media that are inconceivable today. The most obvious of these are virtual and augmented reality services that would strain the networks and phone bills of current 4G users. One panelist mentioned a VR program that requires several gigabytes to download; that’s limited to your home network now. Plus, 5G’s low latency is necessary to prevent motion sickness for VR users.
Imagine if virtual and augmented reality apps became as commonly used as mapping apps are today. That would be a sea change in both how we interact with mobile technology and what we require of our networks. Similarly, wearable technology becomes much more important when there’s a steady stream of useful data about our local environment being fed to us.
Finally, we’ve already seen the start of personal broadcasting, but 5G could make it big news. Current networks are geared for around 90 percent downloading and 10 percent uploading, on the theory that people are more interested in getting than publishing data. 5G is expected to put equal emphasis on upstream and downstream traffic to support autonomous vehicles, remote surgery, and other scenarios where plenty of upload bandwidth is crucial. Consider too the debates about whether police should wear body cameras at all times; 5G could create an environment where anyone could choose to wear one and make it public with a button press or a voice command.
From Predictions to Reality — I am, to use the word from the name of the panel, “stoked” about 5G and its eventual adoption, but I’m less optimistic about how rapidly we’ll see these benefits. First, there’s no agency in charge of making sure that what’s being sold as 5G is truly 5G. Just as some augmented 3G services were marketed as being 4G before 4G was widely available, I expect the same thing to occur with 5G.
The biggest problem isn’t technology but business models. It’s as silly to apply 2017’s network pricing to 5G as it would be to think of our current broadband connections in terms of the AOL “pay by the hour” pricing plan. But some things about pricing models have been extremely sticky. Data caps have rarely been removed from provider plans, and even then, an unlimited data connection seldom takes advantage of the fastest speeds a network can provide. It’s still common for types of data (generic data, streaming video, and tethering, for instance) to be metered at different rates. On my current data plan, a 5G connection could run up bills of $100 per second. The last thing I want is technology that makes me more frightened
about what background apps on my phone can do when I’m not paying attention. Obviously, $100 per second isn’t going to be the pricing model (but don’t be surprised by scattered reports of people with $20,000 phone bills), and it would be difficult for nearly anyone to consume so much data in any real way, but the point stands: we don’t know what 5G’s pricing model will be like.
To me, the true breakthrough would be technology that requires less of my attention, not more, and that means business models without caps or metering. Even more revolutionary would be to eliminate the current necessity of thinking about various tasks as requiring different places. Despite the fact that LTE is faster than many public Wi-Fi hotspots, I still have to think in terms of “what I do at home,” “what I do at Starbucks,” and “what I do when I’m relying on my phone” when I do my Getting Things Done planning. It strikes me as anachronistic that there are things I still have to do at home before heading out for the day and things that I can’t do when on a long trip, but that’s today’s state of affairs. I’d be
more excited about such a change than just about anything else 5G could offer, but I don’t trust the phone companies to give us that.
Instead, you can expect service plans varying based on what the market will bear, even when the technology could enable revolutionary change. The average cellular bill in 2014 was reportedly $73, and I expect 5G plans to try to nudge that number higher. Granted, what you’ll get for that amount will be a higher quality of service along various technological measurements.
What I’d really like to see is a technology that makes broadband speeds truly universal and affordable to all Americans. Here, my bet is on my personal carrier, Google Fi, other scrappy third-party network providers, and to a lesser extent T-Mobile and Sprint, to come up with service prices that change how we think about our monthly phone bill. 5G could help make that possible, but it will be an uphill battle, given how entrenched today’s major carriers are. Nonetheless, cellular network services, alongside frequently monopolistic home Internet services, richly deserve disruption.