The 5G hype is everywhere. It’s telling us that 5G is the best technology ever, and that it will to change the world as we know it. It seems like 5G is going to affect everything we do, and of course make life better. Really?
I know that I’m a natural skeptic. And I’m not the only one. In a recent discussion with a couple colleagues, Bob Landman and Paul Rako, it was clear that we all shared a common feeling. Specifically, is 5G wireless technology really as good as they say? Our doubts are based on valid engineering skepticism and related experience. This discussion actually led to more questions than answers. Questions like: Is 5G really that good? And will 5G really be worth the extra bucks we will be asked to pay?
Anyway, here’s a summary of this discussion:
1. The stated goals of 5G are to increase data rates to 20-Gb/s downlink and 10-Gb/s uplink, decrease latency to 0.5 ms (downlink), and increase subscriber capacity 100 times. These are obviously peak specifications that undoubtedly will rarely be achieved in a normal usage. 1 Gb/s seems doable at times, but imagine thousands (even hundreds or dozens) of users seeking 1-Gb/s service at the same time.
2. The whole 5G concept is based on the small-cell architecture. That’s ok as far as it goes, but to ensure good coverage and high data rates, thousands of these cells are going to be needed for the millimeter-wave (mmWave) bands where useful signals probably won’t travel more that 200-300 meters or so. Even with agile beamforming gain antennas, it will take many cells to cover just an area the size of a city block. And that’s hopefully with good line of sight (LOS) all around. Furthermore, mmWave signals just don’t penetrate walls and other obstacles all that well, if at all, keeping the range super short. Given that an estimated 80% of all calls occur indoors tends to indicate no connection will be made. To make a call, the cell-phone user will have to go outside near a cell.
3. Another issue involves mounting places for all of these cells—are there enough? Only so many utility poles, lamp posts, billboards, and building corners are available. Numerous cities have said no to small cells in some places and approval delays for mounting requests are slowing the installation of 5G in many areas. And should we be paranoid about the potential EMI storms that could develop from so many cells so close together?
4. What about backhaul for all of those thousands of cells? Some have said that backhaul will be via fiber. Maybe, if some fiber is nearby. Will carriers actually invest in digging up streets and sidewalks to add fiber? Probably not. While fiber is a good choice, what if many users try to get 1-Gb/s downloads about the same time. That could tax the cell with data streams of tens to hundreds of gigabits in the fiber. Is that realistic? What may happen is that backhaul will use more mmWave links. That amounts to another complex radio tied to the small-cell radio. Though that’s doable, remember that the backhaul is also going to need its LOS path to work properly.
5. Spectrum is another concern. Is there enough? It doesn’t seem so. The FCC is raking in billions selling spectrum, but is there some for everyone? Some of the spectrum that has been freed up is fragmented narrow slices. The 6-GHz spectrum is promising, but that could be designated as unlicensed. The high data rates and wideband mmWave assignments demand more spectrum and bandwidth. Will it materialize? I doubt that the FCC and others really know what will become available and when. At least they’re trying.
6. Latency. It’s probably not a problem in most use cases. Maybe it’s a limitation in some applications like self-driving cars (autonomous vehicles or AVs) and factory robotic automation. The wireless link has been designed to deliver that small latency. But what happens when the signal goes out into the wired (fiber or UTP) part of the attached network? Oops…
7. Eventually all vehicles, including AVs, will get radios that talk to other nearby vehicles, roadside units, and infrastructure sources to improve safety. The older, already blessed DSRC technology appears to be dead. Instead, the new 5G technology called C-V2X is winning. Hopefully it will work in the 5- to 6-GHz DSRC band and not mmWaves. Will you trust your AV with a short, unreliable mmWave link? A beamforming phased array on the top of your car may help.
8. Will 5G be better than 4G LTE? LTE is very good. Most of the world uses it. LTE delivers reasonable “fast enough” rates for video streaming and the like. Do we really need more speed? Is more speed the point where you can’t tell the difference between 10 Gb/s and 5 Gb/s? The goal is for 5G to replace LTE eventually. Are we ready for that? Many carrier sites haven’t yet upgraded to the faster versions of LTE Advanced and LTE Advanced Pro. The latter upgrade is supposed to get the data rate up to 1 Gb/s. Most of us could live with that. 5G must coexist with LTE on all networks initially. As funding allows expansion, 5G will probably eventually take over the network. The 4G to 5G transition is probably going to take a while.
9. During our discussion, the subject of spectral efficiency came up. The 5G NR specifications say that 30-bps/Hz downlink and 15-bps/Hz uplink are possible. The modulation is some variant of OFDM. But to get to that level of spectral efficiency, you need to use 64QAM or 256QAM on those subcarriers. Good luck with that. With multipath virtually everywhere and the sensitivity of mmWaves to it, those levels of spectral efficiency are a noble goal, but they don’t seem realistic.
10. Antennas could be a problem. 5G relies a great deal on MIMO and phased arrays with agile beamforming. The multiple antenna problem of MIMO is an issue, especially in handsets that already have up to 10 antennas in them, with a maximum of 2 for MIMO. As for phased arrays, the antenna elements themselves are tiny at mmWave frequencies. But arrays of these antennas are much larger, so that they can use hundreds of elements and all of the related circuitry. Will these be too big for small cells?
11. Handsets. Just where are the 5G phones? They are in development. Huawei, Samsung, Xiaomi, LG, and ZTE all have 5G phones ready to go now. The multi-standard, multi-band devices will have to support LTE and 3G as well as 5G. More antennas are needed, and power consumption will probably go up, especially on the mmWave phones. 5G phones will undoubtedly be larger with bigger batteries. And how many cellular band filters will be in those front-ends? Maybe all 5G phones will be like the Lenovo-Motorola Moto Mod that offers a standard LTE smartphone but with a clamp-on 5G accessory if you need it. Whatever the format and content, 5G phones will cost more than the upper-end smartphones today, and your battery may not make it through the day. And we are all wondering what Apple will do. Will a 5G iPhone with a folded color screen be in our future?
12. One thing that seems out of whack is the touting of 5G for the IoT, smart grid, and industrial automation. Most IoT applications only need low speed. Several kbits/s is often enough. The same with electrical utilities and the smart grid. They use wireless but rarely need more than low kbit/s rates for their monitoring and control. Industrial-automation applications are all over the place, but they can usually live with <10-kb/s sensor links and a few Mbits/s on some other equipment. Yes, we know that 5G will work, but it seems like overkill for all of those apps.
13. Will the rural areas of the country ever get 5G? Will it ever get LTE? Some rural towns and villages still have 3G. I’m not making this up. Those folks would be thrilled with 4G LTE. I’m guessing they will not get 5G, as it just too expensive and impractical. There’s no money in it for the carriers. I hate to say it, but is this something the government (FCC?) should deal with? I can’t help but wonder how many people are just underserved.
14. Do mmWaves cause brain cancer? Probably not unless you talk on voice-over-5G for five or more hours per day with the phone against your head. That’s unlikely. But the real answer is we don’t know for sure at this time.
15. Will 5G replace Wi-Fi? Or vice versa? I don’t know who came up with these ideas. If you don’t understand the workings or applications for the various wireless technologies, you may think this is possible. In any case, the answer is a solid no.
16. Will the initial 5G broadband wireless internet access service to be offered by Verizon and others really be competitive against cable or DSL? Most wireless ISPs, and there aren’t many, haven’t done well. They mainly serve the rural and boonie areas with 1 Mb/s or a bit more. If a LOS path can be found, it may work ok. Sprint’s Clearwire service with WiMAX didn’t do all that well. Results will probably be mixed depending on the local environment.
17. As for why you should care, here are a couple of thoughts. The good news is that 5G is a boon to our electronics industry. Business will be good for a while. The bad news is that 5G will simply perpetuate the ongoing preoccupation and addictive damaging obsession with smartphones. Maybe even make it worse.
Lots of questions. We’re not party poopers or negative thinkers. Just a few experienced engineers with opinions and doubts. We trust that the 3GPP developers, equipment manufacturers, and carrier engineers have considered all of this stuff and can reassure us that 5G will be as good as they predict.
If it isn’t as good as they claim, why should we care, since we really can’t do anything about it. As long as LTE remains available, we won’t have to pay the extra they will obviously charge for 5G. In any case, we should be happy and hopeful as my editorial colleague Bill Wong says: “We are going to get 5G anyway.” So, let’s get a more positive attitude, people, and start thinking about 6G right now.