You can’t go anywhere without hearing the term 5G these days, and yet it seems that people are still confused as to what it is and how it will relate to the internet of things. I was at an AT&T event last week where 5G is all the rage, so I figured I’d try to clear some things up for y’all.
The first source of confusion is that when people talk about 5G they may be using the same word to mean different things. The FCC, for example, keeps discussing a 5G auction (it just set the first one for Nov. 14). The agency is referring to several bands of spectrum called millimeter wave. That spectrum is terrible for sending data through glass, trees, rain, etc., but is excellent for providing a lot of capacity. We’ll get back to that in a bit.
Another thing people mean when they talk about 5G is a new standard set forth by the 3GPP, which governs the radio standards used by cellular carriers. This version of 5G can be used on the millimeter wave spectrum or the existing spectrum carriers are currently using for 3G or 4G (also called LTE). The radio standard can also be used to provide wireless connectivity in both mobile situations and for fixed points, where the receiving device stays still. Fixed 5G can be used in factories, in homes, and in large buildings. Mobile 5G is what you would traditionally think of as coming from a cellular carrier.
So what does all this have to do with the internet of things? At the moment, very little. Verizon is deploying fixed 5G to homes as an alternative to fiber broadband service. In that set-up, the 5G signals sent to the home are received by a rooftop device and then converted to in-home Wi-Fi. So your connected lights aren’t getting a 5G upgrade anytime soon.
However, we could see fixed 5G broadband delivered in the field (quite literally, in the case of agriculture) to gateway devices that have access to power. Those gateway devices would likely convert the 5G signal back to something more sensor-friendly, such as traditional cellular, Bluetooth, LoRa, or Wi-Fi. The 5G capacity would be used more as backhaul to get information from the remote networks back to the internet.
In the mobile world, we’ll see 5G IoT devices, but not for a while. First, we’ll see travel modems, and then next year, we’ll see 5G-capable smartphones. Because the 5G standard is so new, the silicon radios used are power hungry. Gordon Mansfield, VP of converged access and device technology at AT&T, says he expects the 5G radios in phones to stay asleep until 5G signals are detected (whether it’s over the millimeter wave spectrum or on another frequency). Once the frequency is detected, those radios will wake up and send data. But for now, mobile 5G feels like a smartphone, automotive, and wired gateway story.
So why the excitement around 5G and IoT? One reason is capacity. When dealing with millimeter wave spectrum, engineers have enormous swaths of spectrum. Mansfield says that unlike traditional cellular channel widths of 10 and 20 megahertz, we could see 400 megahertz channels. Data rates for mobile broadband depend on how many bits you can cram into a single hertz of spectrum; more hertz mean more bits. 5G will also be slightly more spectrally efficient than LTE, which means that each individual hertz will also be able to handle more bits. So it’s a double win.
Unfortunately, this capacity comes at a cost. Namely that, to deal with the short radio waves in this spectrum, carriers have had to build dynamic antennas with a lot of processing power in order to pick up radio waves as they bounce off walls instead of penetrating them. The other option is building a line of sight antennas, which works well across fields, but poorly in urban canyons.
So in urban areas we’re going to see mobile 5G using millimeter wave spectrum rely on reflections outdoors and likely Wi-Fi indoors. The other option, which both AT&T and Verizon will explore, is using spectrum in lower bands, which does penetrate walls. So outdoors, we’ll have the awesome capacity, and indoors, we’ll swap that to Wi-Fi to keep the awesome capacity or else drop to lower-frequency cellular spectrum that doesn’t have as much capacity.
The second, and more interesting component of the 5G radio protocol is that it offers lower latency. It’s also why factories and hospitals are so interested in the technology. At Hannover Messe in April, Ericsson showed off an engine part manufacturing line using 5G that achieved a one-millisecond latency, which allowed for real-time feedback during the milling of the airplane part. This sort of demonstration is cool, but relies on several other processes to be super fast as well. The feedback during manufacturing has to be computed quickly to change the way the part is milled in real time, for example. That’s a function of stream processing and wireless communication.
Still, latency in industrial settings is one place where 5G will make a difference. Again, because of the way millimeter wave spectrum works, we may not see it used inside factories, where large moving metal parts can wreak havoc on wireless networks. But we will see 5G radios using other spectrum bands to allow for more automation without sacrificing latency.
Hopefully you now have a better sense of where 5G will impact the internet of things first, and why. I’d love to continue the discussion if y’all have other ideas or questions.