The scaling of wireless bandwidth is one of the under appreciated technological miracles of the 21st Century.
I wonder: Would even the most ambitious of telecommunications researchers in 1990 have thought it likely that in 2020 we’d be rolling out gigabit wireless speeds to handheld devices over nation-spanning networks?
The scaling of Moore’s Law was clear by then, and certainly fiber optics represented a practical path to near-unlimited wired bandwidth. What were the radio folks thinking at the time?
I still don't really get the point of 5G. It requires so many towers that you're basically building a wired network, but then it requires billions of dollars worth of wireless spectrum in order to get from what amounts to the street in front of your house to inside, even though there are already phone lines and coax going there that could be used. Even for wireless devices, that would get you close enough to use unlicensed spectrum (WiFi) and save billions of dollars worth of licensed spectrum.
The main benefit of cellular is when you're away from WiFi. But unless you're constantly traveling to places without WiFi, you get that from a $20/month plan and the existing cellular network.
5G is not mainly for consumers like you and me. We are not willing to pay significantly more for the faster speeds, most people are plenty happy with 4G. The big difference with the 4G to 5G transition compared to the previous transitions is not speed, but flexibility.
5G aims to allow for completely new applications that were not possible before, and hoping that's where the money will come from. Ultra-reliable and low latency, and massive machine type communications are two new areas 5G is pushing into. The first will allow applications like self driving cars, remote control with immediate feedback, and combined with the increased speeds, augmented and virtual reality. The second area is for the smart city/home type of applications.
This is all from my digital communications professor with ties to Ericsson, but on a personal note I'd be hype to explore distant places in real time by wearing a vr headset and controlling a drone with a 360° camera and <10ms latency. You could be anywhere in the world and actually interact with the environment with just a 5G enabled headset and a 5G-drone renting app. Racing through an abandoned mall! Drone laser tag in a redwood forest!
The benefit is mostly realized in places where you can expect to see the cost savings trickle down and have very high amounts of data for 25$/month. Which, to be honest, is not North America right now, but that certainly will become a reality in some places.
WiFi is rather terrible, and I believe 5G aims to compete with it along with other standards.
I have a 90 sqm two storey house with a detached garage, workshop and quite a bit of land around it and I need 6 WiFi routers to get a signal everywhere, and even then it's spotty and slow.
Sometimes makes me wish I just had a data SIM on all devices and be done with it.
Open source 4g/5G stuff that was posted a while ago looks really nice, but apparently it would be illegal to operate :/
More interesting than 5G is the Amazon's ambitious LEO Project[1], a plan to launch several lower orbit satellites (LEOs) so that plebs can have internet access directly. Will be very useful in dictatorial regimes who censor/control internet in their countries. This will probably render 5G useless.
Probably? There is a fair bit of overlap between cable/fiber and radio in that both are mostly QAM, multichannel based. Radio tends to have a worse SNR and a couple other tricky problems tacked on top.
First of all, this is an excellent explanation of 5G that anyone moderately technical can understand. Great article.
That said, do the numbers from the article concern anyone else at all? I'm generally all for progress, but more than 1 million devices per km2? And massive MIMO antenna arrays spraying EM waves in every part of the spectrum from 600 MHz to 50 GHz?
More than a million devices per km2 ~= 3 million devices per square mile. Am I the only one that thinks that's a bit crazy? I don't want everything from my toaster to my door to my water bottle to be transmitting massive amounts of information to who knows where all the time.
More concerning: the antennas. Current regular MIMO 4G towers are 2x2 or 4x4. Already, there are 5G towers installed today that are 128x128. These are planned to be spaced -extremely- densely in cities, and very close to people (on every floor of offices, on lampposts, etc.) This is a necessity due to the spectrum used. And not only that, but they're using new spectrum that hasn't been used before - instead of the MHz to low GHz range we all know and trust, 5G towers will be up to near 50 GHz (!). Especially the fact that millimeter wave/extremely high frequency is normally blocked by everything from drywall to glass, but now we're intentionally aiming massive amounts of EHF waves with hundreds of antennas in a small space.
Call me a luddite, but I'm more and more thinking that we need a high quality longitudinal study about any effects of this kind of stuff before we go from 2x2/4x4 at well known frequencies to putting 128x128 EHF antenna arrays every hundred feet.
I don't have a study on hand but I expect the effect to be less than lower frequency waves.
First thing to get out of the way, these are not ionizing radiation. Basically, the only thing electromagnetic waves up to visible light can do is heat. UV is borderline.
The question is what they heat. The general idea is that the lower the wavelength, the deeper the penetration. 2/3/4G frequencies will heat your insides, 50 GHz will only heat your skin. Of course the power is so that under normal conditions, the heating is negligible.
As for the effect of millimeter waves, while testing this https://en.wikipedia.org/wiki/Active_Denial_System they subjected people to massive (100kW!) doses of 95 GHz radiation. High enough to actually burn you, and yet, the adverse effects were minimal.
If you are talking about technical problems, like interference, and how to deal with radio waves that are starting to act a bit like light. I guess that 5G is designed by people who know what they are doing. Simply that we are able to make it work is almost like a miracle to me.
Human health would be the first thing on my mind, but there are many other things it could have an effect on - everything from the neighborhood pigeons to trees.
And yes, I am sure that 5G is designed by some brilliant people. Anyone getting that close to the Shannon limit is smarter than me. My hat is off to them.
But the reality is that this stuff just hasn't been trialed in any real way. Sure, the US government found 600 adults to get zapped for a few hours and only injured eight of them. But how about living with it for ten years? What about different frequencies? 100 nm is a world away from 300nm which is a world away from 500nm. Sure, they trialed 95 GHz and it was fine. What about 48 GHz? 25 GHz? All the other frequencies being used? What about with 256 antennae transmitting at higher power than anything before, right next to you, day in and day out for a decade?
One big claim is that all this 5G radiation is blocked by the skin, so we'll all be fine! Oh wait, what about babies? Turns out babies' skin is thin enough that mm waves can penetrate the dermis...
Furthermore, ionizing radiation and heating are not the only two ways EM can harm you. Just look at all the news about blue light for one example! EM radiation in the mm-wave range is still far understudied compared to the cm wave range.
Look, I think that 5G is probably going to be fine, health wise. But for something as big as this, it absolutely deserves at least one good longitudinal study. Of which there have been zero.
Ok, so, for one, you're miss-understanding the MIMO designation, or at least miss-representing it's meaning. 128x128 means that both sender and receiver have 128 antennas. Not that it's an array of 128 * 128 antennas (16384).
Also, the 128x128 5G MIMO deployments I could find were distributed deployments, placing remote radio heads very densely at relevant indoor locations. I assume they're referring to places like large train stations or shopping malls.
Also, regarding the "massive amounts" part: there is little incentive to transmit more than a few hundred mW per channel, just because the mobile device needs to transmit at a similar level anyways, and there are battery and cooling considerations, even neglecting the cost of power amplifiers for these frequencies.
And, the 1 million devices per square kilometer is actually just a dense, standing crowd. It's the average density that can be reached over any sizable area (putting a school class's phones in a box until the end of the lesson would imply some insane density in that box, but there'd be a limit for how many boxes could be arranged into a dense grid).
I understand the MIMO designation, and I am not trying to misrepresent it in my comment. I know that 128x128 is not sixteen thousand (!!!) antennas on one tower, but rather 256, half for rx and half for tx.
That said, two hundred and fifty six antennae on one tower is far, far more (orders of magnitude, plural!) than the typical 2x2 or 4x4 4G towers deployed today. And keep in mind that this is certainly not "just" 256 antennae for one mall, because 5G is blocked by anything thicker than a post-it (the article mentions how future plans include deploying massive MIMO to each floor of a building, as drywall and glass both block many 5G frequencies!)
The reality is that no matter how you want to slice it, 5G means moving from widely spaced 4G towers transmitting at frequencies we have used for around a century to hyper-dense massive deployments of large numbers of antennae transmitting at new and relatively un-studied frequencies. 4G towers are (mostly) kept away from humans getting too close; 5G towers will be on lampposts and on the same floor of your office where someone can touch them.
I'm not saying 5G is inherently bad, or good, I have no idea. I haven't run a longitudinal study, but _neither has anyone else_, and that's what scares me a little. Look, the #1 post on HN right now is how highly trained, expert engineers accidentally made a bridge (one of the oldest human inventions, around for millennia) create an extremely loud eerie noise currently blanketing San Francisco.
And sure, transmit power is low. But it's not just about the power. Today I went out and enjoyed a warming 1000 W/m2 of hundreds of-nm EM radiation all over my body. Then I accidentally looked at the source, which hurt my eyes and disoriented me. Later I played with a device of only 0.01W emitting highly temporally coherent EM radiation. I had to put on safety equipment, because it had the possibility to harm me, including permanently damaging my sight.
So I posted this comment because I wanted to spark a discussion where intelligent people consider the safety implications of 5G, which is very hard to find on the internet. Unfortunately that hasn't happened so far, so I decided to dig a bit more on my own.
Step 1 so far has been Wikipedia, where I found the reassuring sentence "The scientific consensus is that 5G technology is safe.[99][100][101]" Unfortunately, after checking those sources it appears that one is a blog and the other two are news articles about the same event - a press release from the ICNIRP. Further research on the ICNIRP shows that it is a private organization that declared itself that it only considered the thermal dangers of RF in the cm-wave range, and any other dangers do not exist.
And that is the common path to "proving" 5G safety on the internet: "the only way that this could be bad is heating; there's no heating; bam, it's safe, we're done here."
The reality, of course, is that EM waves have many mechanisms of action. So far, we know of the following:
- Ionizing radiation (nuclear weapons, x-rays, etc) can easily kill you or give you cancer
- Non-UV visible light can increase skin aging, cause retinal injuries
- Tons of EM waves can burn you or increase body temperature causing other problems (tissue heating)
- Nerve disturbance from low-frequency RF
That's four methods of ways EM waves can hurt you. Everyone is in agreement about that - no matter how you slice it, there's a minimum of those four. But that's a minimum. Relatively few good quality longitudinal studies have been run about EM waves. "Hey, can we put this weird thing in your home and see if it gives you cancer?" is a hard sell, I guess.
But unfortunately there seems to be not one single high-quality, longitudinal study researching the effects of living with constant mm-wave sources. There are a few short-term studies, but most of these are on the order of hours (!) of exposure.
One interesting thing I did find was a Swiss study showing people living near high voltage power lines were more than two times more likely to get Alzheimer's than people living just 600m away. While that's a completely different type of EM than 5G, it does hint that there may be more harmful mechanisms of EM radiation than we currently know.
Yes, but the "other mechanisms" of EM harm you list do not apply to mm-waves. Imagine being wrapped in lead to protect your organs from harmful mm-waves. That's effectively already the case; mm-waves can't penetrate skin, eyes or anything else.
Moreover, the power levels you're encountering from a base station are minuscule. It's underappreciated as a miracle of engineering that we can reliably communicate with signals that are 10^-12 W. You're at more danger from a 10W fluorescent light bulb.
I'd be more concerned with the fact that the increased quantity and precision of the antennas allows collection of extremely fine-grained location information.
This is an interesting thought. It wouldn't be more precise than the GPS tracker already in your pocket, but it does open up more info to people other than Google.
Also as a note, regular MIMO (not "massive") wouldn't provide location information as the scattering channel is generally considered random.
I have this irrational fear of 5G for some reason and I consider myself a rational person with a pro-technology bias.
This fear stems from a combination of factors leading to unknown unknowns:
1. Little benefit for consumers with 4G (this is admitted even by pro 5G side implementing the tech)
2. Thus leading to question - who benefits when your toaster joins road sensors in the internet of shit.
Webcams already are not doing too hot on the privacy front.
3. Lack of health studies
4. The feeling of being pushed down consumer throats.
5. Last not least the fight between 5G providers/implementers at a state level .
Yeah, I've been slightly concerned about those numbers for a while now. I am in the same boat as you; we think radio waves are totally safe, I get that. But the increase in background spectral density & energy is going to be huge. Makes me feel like we should up our certainty levels a bit more first.
Thanks! I'm glad I'm not the only one. And exactly - it's such a massive jump that it deserves a careful look, which it doesn't seem like it really has gotten so far. I'm not saying it's going to give people cancer, but we just don't know what it may or may not do. Maybe there are absolutely no side effects. Maybe it's even beneficial to health! Hell, exposure to enough EM radiation in the hundreds of nanometers range is actually a requirement for human health.
But maybe it has a negative effect on anything from birds to bacteria to humans, and I think it's worth investigating that before we deploy it extremely rapidly in all of the world's cities.
I mean, the #1 post on HN right now is how highly skilled engineers accidentally made the Golden Gate Bridge emit a loud noise that's blanketing the Bay Area. And that's a bridge, which humans have been building for some 20,000 years. And it's made out of one thing - steel - which we generally understand very well.
I really hope there will be no side effects. But the chance that there will scares me a bit. 5G will be rapidly rolled out at enormous, unprecedented scale and density in nearly all the big world cities - and in 2020 most of the world lives in cities.
One thing to note is that the frequency bands specified only apply to US 5G. Huawei's 5G spectrum uses a much lower frequencies: Around 3.8GHz. This falls in the mid-range spectrum, which makes it cheaper to deploy, because the waves travel further and you need fewer towers. That is the attraction of Huawei, for most countries (outside of the US).
This doesn't make a lot of sense.. There is no "Huawei 5G spectrum". Huawei equipment has to operate in whatever spectrum regulations exist for the country buying the equipment. Countries (including parts of the US) buy Huawei gear because it's cheap, not because there's any special spectrum benefits.
The millimeter wave stuff is in addition to this for downtown cores, stadiums, etc. 5G modulation itself adds max 15-20% throughput compared to 4G LTE so more spectrum is needed to deliver on the weird promises telcos made.
Wifi has been around a long time now. Why can't the hospital devices just use that like everyone else? (Oh except wireless microphones in churches, which are also some special snowflake exceptions for some reason...)
Isn't the ISM bands at 2.4 and 5.8 GHz? That shouldn't be interfered by 3.8 GHz 5G signals. Is it because of out-of-band intermodulation from 5G signals or because there are parts of the 3.8 GHz spectrum used by medical devices not for wireless networking but for sensing/measurements?
Excellent article - still I can't resist a rather large quibble.
The commonly stated assertion that spectrum is a limited resource is not quite accurate.
Spectrum is technically defined as a range of frequencies. That's all. In business parlance spectrum is also attached to large areas of land - this spatial dimension is more important than most people realize.
Radio signals themselves exist in space and time, not just in a range of frequencies. More radio signals - and data bandwidth - can be packed into a given space by shrinking the volume a given signal "occupies".
We can do this by reducing signal power and increasing cell density, in addition to other techniques described in the article. More cells, smaller cells. This is a big part of how 5G expands cell network capacity. But the telecoms have downplayed the effect of this relative to the the claim that spectrum is limited.
The mobile carriers have financial incentives to do this. These incentives are lower costs and monopoly control.
Fewer bigger cells at higher power are cheaper than many smaller cells at lower power. The monopoly part is exclusive use of spectrum on a given piece of land.
The problem is, the legal attachment of spectrum works with very large areas of land (where km^2 is a smallish unit) and large periods of time (years), relative to radio signals. Both attachments are grossly inefficient.
By shrinking cell size (power) and increasing cell density, several orders of magnitude more network bandwidth is possible, plenty even to share (modulo cost of physical infrastructure).
Spectrum scarcity is a myth. The current legal regime enriches monopolists and is otherwise a tremendous waste of potential. We pay higher prices for unnecessarily limited bandwidth.
Spectrum scarcity is also a function of our wasteful utilization.
What you describe (small cells) is one version of what we call spatial multiplexing. Basically, reuse spectrum in physically separate areas. But you can also do this in a way that doesn't sacrifice the centralized, large cell architecture.
Namely via beamforming. You don't have to share spectrum if you're not dumping power in all directions. Already cell towers are split into three sectors; you could continue to increase sectors, or dynamically point beams at individual users. Being suitably isolated from one another, each beam allows spectrum reuse.
It's not free obviously, it requires more expensive base station antennas, but I think a direction we'll be heading in.
Would anyone like to speculate on what new tech we will see that will take advantage of this? I can already stream (to my device) a 4K 60fps video while in a park, and at home I have gigabit fiber (up and down) which is barely ever used to it's full potential. What cool stuff will I be able to do in 10 years time when I have 10Gbps upload while sitting in a park?
It's not clear to me that it makes sense to replace wired with wireless when density is such that fibre can be economically laid to a location.
5G has a lot of possibilities both for improved mobile data and for last mile where it either doesn't exist or is some old 1Mbit ADSL line. But I'm not convinced it will generally make sense to ditch good wired broadband just because 5G is available.
Even if you technically could, I expect the economics won't work out, e.g. throttling/caps/overage charges, for people trying to do a lot of data-heavy things like video streaming if they have an alternative.
I spent over a decade waiting for fiber. Now that it's here, I sure don't want to ditch it. Anyway, it's installed. And the fiber itself can almost certainly be used for higher speeds in the future, now that it's there.
Reminds me a bit of the alleged quote "640K is more memory than anyone will ever need on a computer"
It is difficult to know now, but I think the probability is very high that we will "need" the bandwidth down the line for something we will find useful.
One thing 5G should allow is getting a lot of data point to point wirelessly - specifically for things like oscilloscopes and such, I remember a guy from keysight saying their big push now is make everything work together on the network (which they can now do with things like 5G)
One isn't really supposed to ask these questions, and just assume:
- that every cellphone user in the world is going to just run out tomorrow and buy a phone that's $200-300 more expensive to replace the phones they have now, have been replacing less and less of in the past half a decade, that they barely utilize to its fullest if the network were just prevalent everywhere today
- that some measurable fraction of the towers spaced every 150ft will not be vandalized, broken and sold for scrap in many places in the world and have to be replaced in order to have the throughput as advertised, that the companies will be just excited to replace, whose total cost will be less or the same as current infrastructure costs
- that despite the junk rating of many of the biggest teleco corps, they are well capitalized with pay for such a network from the ever increasing profits they reap from current consumers and wont have to rely on more junk financing and will take great care of the infrastructure just like all the great care they have put into existing infrastructure. Nothing in the future could ever threaten their solvency
Actually, low range of high frequency signal is beneficial. The reason is, that even at relatively large transmission power the signal dissipates quite quickly meaning you can have stations service relatively small area.
This means the spectrum is shared by less users, the uplink is shared by less users, you can serve higher concentration of people.
No mention of unlicensed bands? After the demonstrated superior utilization of those tiny portions of this range allocated for ISM? My goodness, it's almost as if FCC works only to perpetuate outdated "giant telco" models to the detriment of all consumers!
Out here in the country we'll operate our own "small cells" without permission from ATTVZN and without paying FCC a cent. If anyone notices there will be "investigations" but mostly no one will notice because physics. Eventually industrial users of this tech will realize "hey we don't need those telco goofballs either!" and their lobbyists will muscle through some exceptions. Eventually everyone whose house has sheetrock walls will be so excepted.
This article wasn't about unlicensed/ISM radio, though there have certainly been advances there as well. I'm genuinely curious about which parts of the ISM range you see with "superior" utilization compared to cellular networks - I work with those type of radios for a living.
For most of the spectrum, the devices and services that consumers may purchase are limited to those available from giant hierarchical oligopolies. In the ISM bands, we have a bunch of different technologies, but most importantly Wifi. There are much less stringent barriers to anyone using or selling devices and services that use Wifi. And it all works! Those who live in dense urban areas sometimes scoff at Wifi, but what they're seeing is the tiny segments of spectrum allocated to unlicensed use, not inherent problems with open spectrum. The excuse offered in the past for spectrum oligopolies was "oh it's physics" because the very simplest radio technology we could imagine in the 1930s does suffer from interference at lower frequencies. Of course we've had more advanced radio tech for a long time, but ignore that... at the frequencies under discussion for 5G, interference over any but the shortest range is no longer a threat. That means that most of the claimed justification for telecom monopolies (and perhaps the agency created to enforce those monopolies) goes away. It's sad that TFA is still focused on "carriers".
Readit News
I wonder: Would even the most ambitious of telecommunications researchers in 1990 have thought it likely that in 2020 we’d be rolling out gigabit wireless speeds to handheld devices over nation-spanning networks?
The scaling of Moore’s Law was clear by then, and certainly fiber optics represented a practical path to near-unlimited wired bandwidth. What were the radio folks thinking at the time?
The main benefit of cellular is when you're away from WiFi. But unless you're constantly traveling to places without WiFi, you get that from a $20/month plan and the existing cellular network.
5G aims to allow for completely new applications that were not possible before, and hoping that's where the money will come from. Ultra-reliable and low latency, and massive machine type communications are two new areas 5G is pushing into. The first will allow applications like self driving cars, remote control with immediate feedback, and combined with the increased speeds, augmented and virtual reality. The second area is for the smart city/home type of applications.
This is all from my digital communications professor with ties to Ericsson, but on a personal note I'd be hype to explore distant places in real time by wearing a vr headset and controlling a drone with a 360° camera and <10ms latency. You could be anywhere in the world and actually interact with the environment with just a 5G enabled headset and a 5G-drone renting app. Racing through an abandoned mall! Drone laser tag in a redwood forest!
>The main benefit of cellular is when you're away from WiFi.
For you maybe.
The billions you mention are a result of government auctioning, it doesn't mean anything to users besides added cost.
I have a 90 sqm two storey house with a detached garage, workshop and quite a bit of land around it and I need 6 WiFi routers to get a signal everywhere, and even then it's spotty and slow.
Sometimes makes me wish I just had a data SIM on all devices and be done with it.
Open source 4g/5G stuff that was posted a while ago looks really nice, but apparently it would be illegal to operate :/
Edit: apparently 5G does not use UWB
[1] https://www.theverge.com/2019/4/4/18295310/amazon-project-ku...
That said, do the numbers from the article concern anyone else at all? I'm generally all for progress, but more than 1 million devices per km2? And massive MIMO antenna arrays spraying EM waves in every part of the spectrum from 600 MHz to 50 GHz?
More than a million devices per km2 ~= 3 million devices per square mile. Am I the only one that thinks that's a bit crazy? I don't want everything from my toaster to my door to my water bottle to be transmitting massive amounts of information to who knows where all the time.
More concerning: the antennas. Current regular MIMO 4G towers are 2x2 or 4x4. Already, there are 5G towers installed today that are 128x128. These are planned to be spaced -extremely- densely in cities, and very close to people (on every floor of offices, on lampposts, etc.) This is a necessity due to the spectrum used. And not only that, but they're using new spectrum that hasn't been used before - instead of the MHz to low GHz range we all know and trust, 5G towers will be up to near 50 GHz (!). Especially the fact that millimeter wave/extremely high frequency is normally blocked by everything from drywall to glass, but now we're intentionally aiming massive amounts of EHF waves with hundreds of antennas in a small space.
Call me a luddite, but I'm more and more thinking that we need a high quality longitudinal study about any effects of this kind of stuff before we go from 2x2/4x4 at well known frequencies to putting 128x128 EHF antenna arrays every hundred feet.
I don't have a study on hand but I expect the effect to be less than lower frequency waves.
First thing to get out of the way, these are not ionizing radiation. Basically, the only thing electromagnetic waves up to visible light can do is heat. UV is borderline.
The question is what they heat. The general idea is that the lower the wavelength, the deeper the penetration. 2/3/4G frequencies will heat your insides, 50 GHz will only heat your skin. Of course the power is so that under normal conditions, the heating is negligible.
As for the effect of millimeter waves, while testing this https://en.wikipedia.org/wiki/Active_Denial_System they subjected people to massive (100kW!) doses of 95 GHz radiation. High enough to actually burn you, and yet, the adverse effects were minimal.
If you are talking about technical problems, like interference, and how to deal with radio waves that are starting to act a bit like light. I guess that 5G is designed by people who know what they are doing. Simply that we are able to make it work is almost like a miracle to me.
And yes, I am sure that 5G is designed by some brilliant people. Anyone getting that close to the Shannon limit is smarter than me. My hat is off to them.
But the reality is that this stuff just hasn't been trialed in any real way. Sure, the US government found 600 adults to get zapped for a few hours and only injured eight of them. But how about living with it for ten years? What about different frequencies? 100 nm is a world away from 300nm which is a world away from 500nm. Sure, they trialed 95 GHz and it was fine. What about 48 GHz? 25 GHz? All the other frequencies being used? What about with 256 antennae transmitting at higher power than anything before, right next to you, day in and day out for a decade?
One big claim is that all this 5G radiation is blocked by the skin, so we'll all be fine! Oh wait, what about babies? Turns out babies' skin is thin enough that mm waves can penetrate the dermis...
Furthermore, ionizing radiation and heating are not the only two ways EM can harm you. Just look at all the news about blue light for one example! EM radiation in the mm-wave range is still far understudied compared to the cm wave range.
Look, I think that 5G is probably going to be fine, health wise. But for something as big as this, it absolutely deserves at least one good longitudinal study. Of which there have been zero.
And yet a domestic microwave at 800W is enough to both burn the skin in minutes and cause nerve cell damage in few seconds.
Also, the 128x128 5G MIMO deployments I could find were distributed deployments, placing remote radio heads very densely at relevant indoor locations. I assume they're referring to places like large train stations or shopping malls.
Also, regarding the "massive amounts" part: there is little incentive to transmit more than a few hundred mW per channel, just because the mobile device needs to transmit at a similar level anyways, and there are battery and cooling considerations, even neglecting the cost of power amplifiers for these frequencies.
And, the 1 million devices per square kilometer is actually just a dense, standing crowd. It's the average density that can be reached over any sizable area (putting a school class's phones in a box until the end of the lesson would imply some insane density in that box, but there'd be a limit for how many boxes could be arranged into a dense grid).
I understand the MIMO designation, and I am not trying to misrepresent it in my comment. I know that 128x128 is not sixteen thousand (!!!) antennas on one tower, but rather 256, half for rx and half for tx.
That said, two hundred and fifty six antennae on one tower is far, far more (orders of magnitude, plural!) than the typical 2x2 or 4x4 4G towers deployed today. And keep in mind that this is certainly not "just" 256 antennae for one mall, because 5G is blocked by anything thicker than a post-it (the article mentions how future plans include deploying massive MIMO to each floor of a building, as drywall and glass both block many 5G frequencies!)
The reality is that no matter how you want to slice it, 5G means moving from widely spaced 4G towers transmitting at frequencies we have used for around a century to hyper-dense massive deployments of large numbers of antennae transmitting at new and relatively un-studied frequencies. 4G towers are (mostly) kept away from humans getting too close; 5G towers will be on lampposts and on the same floor of your office where someone can touch them.
I'm not saying 5G is inherently bad, or good, I have no idea. I haven't run a longitudinal study, but _neither has anyone else_, and that's what scares me a little. Look, the #1 post on HN right now is how highly trained, expert engineers accidentally made a bridge (one of the oldest human inventions, around for millennia) create an extremely loud eerie noise currently blanketing San Francisco.
And sure, transmit power is low. But it's not just about the power. Today I went out and enjoyed a warming 1000 W/m2 of hundreds of-nm EM radiation all over my body. Then I accidentally looked at the source, which hurt my eyes and disoriented me. Later I played with a device of only 0.01W emitting highly temporally coherent EM radiation. I had to put on safety equipment, because it had the possibility to harm me, including permanently damaging my sight.
Step 1 so far has been Wikipedia, where I found the reassuring sentence "The scientific consensus is that 5G technology is safe.[99][100][101]" Unfortunately, after checking those sources it appears that one is a blog and the other two are news articles about the same event - a press release from the ICNIRP. Further research on the ICNIRP shows that it is a private organization that declared itself that it only considered the thermal dangers of RF in the cm-wave range, and any other dangers do not exist.
And that is the common path to "proving" 5G safety on the internet: "the only way that this could be bad is heating; there's no heating; bam, it's safe, we're done here."
The reality, of course, is that EM waves have many mechanisms of action. So far, we know of the following:
- Ionizing radiation (nuclear weapons, x-rays, etc) can easily kill you or give you cancer
- Non-UV visible light can increase skin aging, cause retinal injuries
- Tons of EM waves can burn you or increase body temperature causing other problems (tissue heating)
- Nerve disturbance from low-frequency RF
That's four methods of ways EM waves can hurt you. Everyone is in agreement about that - no matter how you slice it, there's a minimum of those four. But that's a minimum. Relatively few good quality longitudinal studies have been run about EM waves. "Hey, can we put this weird thing in your home and see if it gives you cancer?" is a hard sell, I guess.
But unfortunately there seems to be not one single high-quality, longitudinal study researching the effects of living with constant mm-wave sources. There are a few short-term studies, but most of these are on the order of hours (!) of exposure.
One interesting thing I did find was a Swiss study showing people living near high voltage power lines were more than two times more likely to get Alzheimer's than people living just 600m away. While that's a completely different type of EM than 5G, it does hint that there may be more harmful mechanisms of EM radiation than we currently know.
Moreover, the power levels you're encountering from a base station are minuscule. It's underappreciated as a miracle of engineering that we can reliably communicate with signals that are 10^-12 W. You're at more danger from a 10W fluorescent light bulb.
Also as a note, regular MIMO (not "massive") wouldn't provide location information as the scattering channel is generally considered random.
This fear stems from a combination of factors leading to unknown unknowns:
1. Little benefit for consumers with 4G (this is admitted even by pro 5G side implementing the tech)
2. Thus leading to question - who benefits when your toaster joins road sensors in the internet of shit. Webcams already are not doing too hot on the privacy front.
3. Lack of health studies
4. The feeling of being pushed down consumer throats.
5. Last not least the fight between 5G providers/implementers at a state level .
Why not at least do some health studies?
Dead Comment
But maybe it has a negative effect on anything from birds to bacteria to humans, and I think it's worth investigating that before we deploy it extremely rapidly in all of the world's cities.
I mean, the #1 post on HN right now is how highly skilled engineers accidentally made the Golden Gate Bridge emit a loud noise that's blanketing the Bay Area. And that's a bridge, which humans have been building for some 20,000 years. And it's made out of one thing - steel - which we generally understand very well.
I really hope there will be no side effects. But the chance that there will scares me a bit. 5G will be rapidly rolled out at enormous, unprecedented scale and density in nearly all the big world cities - and in 2020 most of the world lives in cities.
A really big increase in a number without a framework to contextual it doesn't mean anything.
Ref: https://www.huawei.com/en/about-huawei/public-policy/5g-spec....
EDIT: Fixed typos.
https://www.fcc.gov/document/fcc-expands-flexible-use-c-band... "Makes 280 Megahertz of the 3.7-4.2 GHz Band Available for 5G Services While Relocating Existing Satellite Operations to the Upper Part of the Band"
The millimeter wave stuff is in addition to this for downtown cores, stadiums, etc. 5G modulation itself adds max 15-20% throughput compared to 4G LTE so more spectrum is needed to deliver on the weird promises telcos made.
The commonly stated assertion that spectrum is a limited resource is not quite accurate.
Spectrum is technically defined as a range of frequencies. That's all. In business parlance spectrum is also attached to large areas of land - this spatial dimension is more important than most people realize.
Radio signals themselves exist in space and time, not just in a range of frequencies. More radio signals - and data bandwidth - can be packed into a given space by shrinking the volume a given signal "occupies".
We can do this by reducing signal power and increasing cell density, in addition to other techniques described in the article. More cells, smaller cells. This is a big part of how 5G expands cell network capacity. But the telecoms have downplayed the effect of this relative to the the claim that spectrum is limited.
The mobile carriers have financial incentives to do this. These incentives are lower costs and monopoly control. Fewer bigger cells at higher power are cheaper than many smaller cells at lower power. The monopoly part is exclusive use of spectrum on a given piece of land.
The problem is, the legal attachment of spectrum works with very large areas of land (where km^2 is a smallish unit) and large periods of time (years), relative to radio signals. Both attachments are grossly inefficient.
By shrinking cell size (power) and increasing cell density, several orders of magnitude more network bandwidth is possible, plenty even to share (modulo cost of physical infrastructure).
Spectrum scarcity is a myth. The current legal regime enriches monopolists and is otherwise a tremendous waste of potential. We pay higher prices for unnecessarily limited bandwidth.
What you describe (small cells) is one version of what we call spatial multiplexing. Basically, reuse spectrum in physically separate areas. But you can also do this in a way that doesn't sacrifice the centralized, large cell architecture.
Namely via beamforming. You don't have to share spectrum if you're not dumping power in all directions. Already cell towers are split into three sectors; you could continue to increase sectors, or dynamically point beams at individual users. Being suitably isolated from one another, each beam allows spectrum reuse.
It's not free obviously, it requires more expensive base station antennas, but I think a direction we'll be heading in.
You'll have lower energy use modes available for battery powered IoT devices.
You'll be able to be in a very dense crowd and have your services still work.
It is also lower latency that 4G, so things like gaming over your 5G connection become more possible.
These may not matter to you, but they are use-cases that 5G allows.
5G has a lot of possibilities both for improved mobile data and for last mile where it either doesn't exist or is some old 1Mbit ADSL line. But I'm not convinced it will generally make sense to ditch good wired broadband just because 5G is available.
Even if you technically could, I expect the economics won't work out, e.g. throttling/caps/overage charges, for people trying to do a lot of data-heavy things like video streaming if they have an alternative.
It is difficult to know now, but I think the probability is very high that we will "need" the bandwidth down the line for something we will find useful.
- that every cellphone user in the world is going to just run out tomorrow and buy a phone that's $200-300 more expensive to replace the phones they have now, have been replacing less and less of in the past half a decade, that they barely utilize to its fullest if the network were just prevalent everywhere today
- that some measurable fraction of the towers spaced every 150ft will not be vandalized, broken and sold for scrap in many places in the world and have to be replaced in order to have the throughput as advertised, that the companies will be just excited to replace, whose total cost will be less or the same as current infrastructure costs
- that despite the junk rating of many of the biggest teleco corps, they are well capitalized with pay for such a network from the ever increasing profits they reap from current consumers and wont have to rely on more junk financing and will take great care of the infrastructure just like all the great care they have put into existing infrastructure. Nothing in the future could ever threaten their solvency
I wish them all the best!
This means the spectrum is shared by less users, the uplink is shared by less users, you can serve higher concentration of people.
Out here in the country we'll operate our own "small cells" without permission from ATTVZN and without paying FCC a cent. If anyone notices there will be "investigations" but mostly no one will notice because physics. Eventually industrial users of this tech will realize "hey we don't need those telco goofballs either!" and their lobbyists will muscle through some exceptions. Eventually everyone whose house has sheetrock walls will be so excepted.
I had the hope, 5G would finally solve the coverage problems in rural areas. sounds like it's just a vanity project after all...