5G Skeptic | Hacker News Reader

5G Skeptic

tbray.org

191 pointszdw253 comments

One of the primary reasons that carriers are interested in 5G is because for the last decade they have been swapping spit (trading customers between each other, carrier A lowers the prices and take customers from carrier B, who responds in kind and takes from carrier C who responds with a marketing campaign with some deal on a phone etc. and etc.)

What they all are craving is new revenue. Once everyone who wants a cell phone, has one, it becomes a commodity item and service and that means, if anything, one price war after another. This is why they are interested in 5G and MEC (Mobile Edge Cloud) they are looking for new revenue. Right now, as part of the 5G push is the idea of wireless home internet, which is basically just a wifi AP that connects to the internet via the wireless network. To do that of course, you need a lot more capacity, 5G offers that to some extent, but realistically you need a ton more spectrum, which is why you have seen a spate of auctions lately.

Make no mistake here, 5G Home internet is most certainly not a cost savings approach for the carriers, they are literally spending billions on it, in hopes of luring customers away from the cable cos and telcos or other wired providers. Will it work? Perhaps. Time will tell. But to be sure, in my view, this is very good for the consumer of internet services, it will almost assuredly provide the customer with a lower cost per Mbps due to increased competition particularly with all the LEOS (Low Earth Orbit Sats) also in the mix.

> But to be sure, in my view, this is very good for the consumer of internet services

This is one of the things I'm excited about. Even if 5G home internet won't lure you away from your wired connection, it will lure enough customers away that your ISP will need to treat you better (for fear that you might be lured away).

Lots of people with Verizon and AT&T would say that they would never switch to T-Mobile in the 2013-2020 time-range. Still, T-Mobile lured enough customers away that Verizon and AT&T started treating their customers a lot better.

Yes, Tmo is catching up with ATT and VZ on quality, the main reason they bought Sprint (remember them) is because Sprint had a shit ton of spectrum but they weren't building it out because they were capital starved (for a bunch of reasons). So now, TMo has the most low and medium band spectrum and they are catching up on build out. Vz has the most spectrum if you include the mm-wave, but it's difficult to build out.

All the bandwidth competition is a good thing for the consumer and hopefully some cool stuff gets developed to take advantage of it.

I prefer my fiber connection to a 5G connection. My fear is that they will do the math and find they can't compete (I don't know if this is actually the case). It could be the end of physical connections.

Me too, nothing beats fiber if you can get it. I used to have Fios where I used to live and nothing beats it, but where I live now, you cannot get any fiber to the home, but I can get the mm-wave 5G and that is a pretty close second, the speed on my mm-wave is actually faster than the Fios I had before but now there is NGPON2 which is the new tech behind it and that can actually get you to 10Gbps. Fiber will always be faster most likely.

A lot of people don't have fiber options to begin with (especially rural users). I wonder if 5G/NR will give them better service than 4G/LTE.

I doubt wireless carriers will ever be able to seriously compete with built out fiber in the long run. Once it is there, the costs have been paid and the provider is mainly just raking in residuals while keeping the lights on.

Rural user here - yes, 5G is an absolute godsend, 4G was over congested and had extremely unreliable ping (the other option was 2mbit ADSL), I now have a fairly stable 250/100 Internet connection, with much lower latency

> A lot of people don't have fiber options to begin with (especially rural users). I wonder if 5G/NR will give them better service than 4G/LTE.

The flip side to that is the "many rural users are at the edges of existing coverage networks".

A 4G network can reach about 10 miles (16km) - a 5G network is about 1000 feet or 0.2 miles (about 300 meters).

Unless the providers are placing these on every other power pole, most rural users aren't likely to see 5G coverage. For what it's worth, my parents' house is about 0.1 miles from the road and 0.5 miles from the next nearest neighbor.

I believe that it is unlikely that wireless 5G will get out there (they don't have 4G service either - they're in a valley and use a femtocell for home phone use).

They don't have cable or fiber either and switched from a load balanced pair of DSL lines to Starlink.

I thought 5G in that 600mhz band has much better coverage. Isn't it only ultra wideband that's that low?

(searching) - yes it can...

https://venturebeat.com/2019/12/10/the-definitive-guide-to-5...

> One low band (600-700MHz) tower can cover hundreds of square miles with 5G service that ranges in speed from 30 to 250 megabits per second (Mbps).

Though... I'm still going to be skeptical with the "there are still some significant areas that lack 4g coverage"

https://fcc.maps.arcgis.com/apps/webappviewer/index.html?id=...

Look at the "how many areas only have one provider" or that you can see the topography and valleys in the coverage map.

The basics are this, the lower the frequency because lower frequencies propagate better (https://en.wikipedia.org/wiki/Free-space_path_loss)

but...

you will get less capacity, because at the lower frequencies there is less spectrum (it is simple arithmetic there is 1000Mhz of spectrum below 1Ghz, but there is 100Ghz of spectrum below 100Ghz (setting aside the details about how it gets allocated etc.)

The amount of capacity you can achieve is based on Shannon limit https://en.wikipedia.org/wiki/Shannon%E2%80%93Hartley_theore...

it's the classic engineering trade off, better propagation vs. more capacity.

that is why the carriers want low, medium and high band in their inventory,

low is considered < 1Ghz, medium is considered 1Ghz - 10Ghz, high is > 10Ghz, roughly speaking

probably make sense in rural areas, fewer users, larger area.

I can’t speak for anywhere else, but in Chicago Verizon is literally placing 5G UWB bases on every other light pole.

What will the coverage be like outside of Freeport, IL though?

Its this type of poles that I'm more interested in - https://www.google.com/maps/@42.2639251,-89.787017,3a,75y,65...

Surely you can recognize the difference between that and somewhere rural enough that they don't even have lightpoles and if they did, a 5g cell on one could reach at most one customer anyways.

Not just rural. I'm in downtown Chicago and the cheapest wired connection I can get to my home is $71,000 install and $800/month from Comcast.

Finally got T-Mobile Home Internet (5G). Shipped the access point overnight to me. Plugged in. 600Mbps peak. Never seen a slowdown that affects my heavyweight Net use (e.g. streaming HD).

lol what part of downtown Chicago is that per chance? Comcast is required by law to serve every household in their franchise area so I’m extremely curious.

Yeah ... I think I have 3 choices for gigabit internet in my building, and it's probably one of the smallest unit count buildings in the loop.

The key difference is that wireless is completely federally regulated, and the FCC doesn’t hold them to the same rules as wired providers. They don’t have the same net neutrality requirements and can and do rate limit based on a variety of factors.

For 5G, home wireless is bottom of the line and gets de-prioritized behind phone customers.

The playbook is pretty obvious based on Verizon’s behavior. They’ve halted most fiber deployments in urban areas in favor of mm-wave 5G. The FCC declared mm-wave a national defense priority and it is not subject to any local regulation.

Spectrum and Comcast suck, but they are the lesser evil compared to telcos imo.

Both wired and wireless are regulated by the FCC, neither have net neutrality restrictions anymore, the went bye bye in the U.S. at least, a couple years ago and I seriously doubt it will come back ever.

> The FCC declared mm-wave a national defense priority and it is not subject to any local regulation.

I am not sure where you are getting that but mm-wave is absolutely still regulated locally just like anything else. The main difference is the size of the radios for mm-wave are small and go on poles and the FCC did set a price ceiling on what utilities can charge for pole space. But there are also low and mid-band small cells that go on poles as well. But the wireless companies absolutely still have to get local permits to install any wireless system on poles or pretty much anything.

The permitting process is essentially, apply for a permit, the city has 30 days to make aesthetic suggestions and cannot deny the permit.

In my city they are placed in all sorts of bizarre locations that the zoning board would not approve such as, in the middle of lawns, within 8 feet of an existing pole owned by a competing utility, in the middle of a sidewalk, etc.

There is one silver lining: mmWave 5G has a much shorter range and therefore reduced light pollution compared with the spectrum of 4G. This is good collectively good for users of the network.

My neighborhood is less than 15 years old and CenturyLink has no plans to run fiber to my neighborhood. New subdivisions being built are including fiber.

The 5G home internet options can be good for people in older homes or neighborhoods or even rural area which have garbage internet.

> New subdivisions being built are including fiber.

This is good to hear, as that was my main concern. That the deployment of fiber would freeze in the state it was in 2020.

5g home internet probably isn't as good as fiber, but so far with T-Mo‘s $50/mo plan I average 250/30mbit with around 30ms latency. In a relatively dense newer suburb, Att only offers 50/10 "advertised speed" DSL and the cable company's 100/15 is over $70 plus rental fees. The biggest drawback of 5g is CG-NAT which requires tunneling if you want to host web services.

> it will lure enough customers away that your ISP will need to treat you better (for fear that you might be lured away).

I'm worried that the end-game is that you'll still be stuck with equally-shitty providers, just that instead of having to take one turd you'll be able to pick between two equally-smelly ones.

It's not like Comcast support or billing is any better in areas where they have competition, and outside of rare areas where there are local competitors (municipal broadband, etc), Comcast's competition in the form of Verizon/Cox/Spectrum isn't any better and has exactly the same flaws.

The only thing they can potentially compete on is price, but my understanding is that price is never really the problem, it's all the hidden costs such as surcharges/billing issues, technical issues where tech support is horrible and you end up with no internet for days/weeks, etc - something you can't predict in advance when choosing a provider.

As someone who's dealt with a number of providers across multiple countries, my takeaway is that the entire telecoms industry is rotten and the only solution is to expect them to be shit, plan in advance for when they screw you over (such as maintaining a backup connection) and not give them any leverage (no long-term contracts) so that you can just walk away when they become a problem.

If your ISP is AT&T or Verizon, switching to 5G is not switching ISPs.

So in short, 5G isn't particularly better than 4G all else being equal, but it is a prerequisite to expanding and improving service in ways that matter.

Does that sound right?

Basically, 5G introduces some new technology that increases the so called spectral efficiency (bps/hz) that is how much tput you can deliver over the allocated spectrum. 5G introduces things like mm-wave tech using phased array antennas, better massive MIMO etc.

But even with the tech, it's probably not enough to handle the demand they are hoping for, hence they spectrum buying spree they have been on lately.

It goes back to the Shannon law talked about here, you either have to build more cells to make the signal better (better signal lets you pack more bps per hz, or you need more spectrum. Both are expensive, and most of them are doing both.

My 5G connection is better (faster, more reliable) than my fixed line connection

and only £20/month more for 10x download, 10x upload speed of the fixed line connnextion

Genuinely, how's the ping, could you play FPS games? Are you including the overhead of paying the line rental?

VSDL2 — Plusnet £35/month 39mbps down, 8mbps up, 30-300ms RTT (it can be very variable depending on time of day)

5G — Vodafone £50/month ex VAT, 350-390mbps down, 40mbps up, 30ms RTT (and stable)

Cell tower occasionally goes offline but in general the Vodafone connection is way better than the fixed line one

I don't know the details of why but ping times on residential connections in the UK seem to run somewhat higher than those I've experienced in the US. I have a cable connection of over 500Mbps and the ping time is ~20ms to both my nearest speed test location (about 50 miles away) and 8.8.8.8. 3ms is within my home network. 15ms is to my first hop at my ISP(!!).. then 2-3ms is the rest. My DSL connection is somewhat worse but with a similar profile.

I think it’s a factor of how the networks are constructed and how ISP backhaul works

As far as I can remember anyone on BT (or rebranded BT) gets dumped into their backbone that then exits onto the internet in a couple of places but it’s a while since I looked at it

Is your experience of US connections based on connections in large metro areas, rather than rural ones?

In recent years, yes. Last time I used the Internet in rural US was dialup(!)

Looking at a Twitter search for people's 'Speedtest' results, though, provides anecdata. People in all sorts of countries seem to enjoy sub 10ms on even modest connections (e.g. https://twitter.com/grailph/status/1507258781574778883), whereas in the UK even a "good" connection gets a 10-15ms lag baked in.

Have you tried EE's 5G? Their unlimited business plan costs £30/month and at least in London they usually have better speeds and coverage.

I'm currently testing Smarty (Three sub brand). £16/month, unlimited data, 30 days contract. Just outside my front door: https://www.celsoazevedo.com/files/random/Screenshot_2022032...

But I live in the basement of an old building... signal sucks inside, so I'm stuck with a 65/15Mbps FTTC connection (also with PlusNet).

Vodafone is the only 5G tower within range - it’s about 500m away other towers are at least a mile and haven’t been upgraded yet

Funnily when we move to our current address the VF tower was a 2G infill so we switched to Three for better coverage and throughout

> Right now, as part of the 5G push is the idea of wireless home internet, which is basically just a wifi AP that connects to the internet via the wireless network, [...] in hopes of luring customers away from the cable cos and telcos or other wired providers.

I'm not saying they're not trying to create new demand, but this particular argument may be dependent on local context.

Over here in France, there's also a strong campaign for 5G. But in the consumer market, wired internet providers are the same companies that provide cell phone service. This may help in low-density areas, although there's also a big push from government to deploy fiber even in rural areas.

You make an interesting point because here Verizon and ATT are local telcos as well so they can potentially cannibalize some of their own customer based in their serving areas, but both have had wired broadband service options in their serving areas which as stalled in terms of growth over the past few years. So I think they are looking at wireless, which is a cash cow for them, and probably figure they may win more customers from the cableco in their serving areas with a wireless option that does not require any installation or retrofit work at their homes. Bringing fiber into the home can be expensive, and DSL (copper) will not get you the speeds people are looking for these days for broadband service.

You are right, this is all a very U.S. skewed view of things. I do not know the European market but my guess is those providers are probably also looking for some revenue boost from 5G.

> DSL (copper) will not get you the speeds people are looking for these days for broadband service

The German Telekom (same company running T-Mobile, coincidentally) has been surprisingly good at milking their existing crappy copper lines using vectoring. They run fiber up to the distribution boxes on the side of the road (up to a few hundred meters away from houses) and vector their way from there with speeds of up to 250 Mbit/s (which is great for German standards and honestly I'm okay with less).

Maybe they'll be able to increase the rates with 5G or save on distribution cost, but I'm not so sure about either. The distribution boxes are quite large and power hungry with serious cooling nowadays, but they're probably still below the complexity and cost of a 5G tower.

Is that rural areas?

My parents live in a rural-ish town, and have FTTH. The connection from the distribution box up the road is made on an ad-hoc basis. The technician pulls the fiber on the posts used for electricity delivery. It's only been available for less than two years, though.

Sure, if you have no posts and have to dig up the sidewalk or something, it would be different.

But the major advantage of this approach is that distribution boxes are passive, just connectors, so they should be pretty cheap to install and operate.

Here in Japan, FTTH coverage is pretty good but there are many condos that can't sign up any FTTH due to its facility, but provide own crappy internet. Major telco companies are going to deploy 5G to provide better service to such condos.

> But to be sure, in my view, this is very good for the consumer of internet services, it will almost assuredly provide the customer with a lower cost per Mbps due to increased competition particularly with all the LEOS (Low Earth Orbit Sats) also in the mix.

The problem is that all these services improve download--which doesn't matter all that much.

Technology, however, is driven by upload. Every time upload speeds jumped by an order of magnitude, we saw a whole bunch of new computers applications.

Sadly, computers have been knecapped by a maximum 10M upload speed for almost 20 years now. NAT was just an extra gunshot wound to add to the misery.

> [...] in hopes of luring customers away from the cable cos and telcos or other wired providers.

For some, like me it's not a mere lure but a liberation - out of sheer bad luck, the fixed line internet speed at the last 3 places I've moved in a UK city have gotten slower each time, ending up at <2Mbit in the centre. There is no fiber, if you have fiber know that you are merely lucky, for everyone else it's a question of how noisy your copper twisted pair is, which will only get worse.

LTE internet is my only practical option, and it's not simply competitive, it completely obliterates the fixed line competition in my area providing me with 30-100Mbit (depends on time of day), for £35/mo no contract.

I know this is different in the US, but right now in the UK there are 3 major LTE providers which all have unlimited data plans! - even better, 2 of them have unlimited plans without a contract, which IMO is vital because you must test out reception. LTE definitely beats ADSL in this country, only fiber can compete, but availability is very patchy.

> £35/mo

And you can find cheaper. For example, Smarty (Three sub brand), £16/month, unlimited data, 30 days contract. When on 5G (their 4G sucks), I get up to 800Mbps up and 100Mbps down: https://www.celsoazevedo.com/files/random/Screenshot_2022032...

I tried them first for quite a while, but they have a high contention ratio, that's why it's so cheap... although it may depend on where you are. I could pull 100Mbit in the middle of the night but it would go down to 3Mbit in the middle of the day (you can burst higher but the average even over 20 seconds was very low and not very usable).

I'm using it for work so 3Mbit in the middle of the day didn't cut it, vodafone/voxi gives me 30Mbit at that time so well worth the extra money. Only issue I have with vodafone is some weird MTU stuff going on. These LTE modems have some kind of end point MTU negotiation which is quite low on Three (1300 ish), on vodafone it starts on 1500 but then later the effective mtu changes to 1464 which is kinda maddening because the modem doesn't seem to detect this and your larger outbound packets will just vanish (doesn't even send back a fragmentation packet)... so now I have to manually set all the wifi interfaces to 1464, which is pretty silly (also silly that consumer LTE routers don't let you set this, but to be fair it's highly variable between networks and it makes sense that most people shouldn't be playing with it - if only it worked properly)... anyway other than that little issue it's pretty damn smooth.

Oh also I tried EE... which I recommend avoiding, I will never try them again unless they release a no-contract service, they couldn't even activate my sim, customer service is a nightmare, failed promise after failed promise, took me 6months to be rid of them and didn't get all my money back after they didn't deliver even 1 byte. Was a good reminder of why not to do contracts where possible.

I think these issues will work themselves out as more and more people start using this stuff for home broadband.

We tried EE 5G in a bigish UK town and it was a big fail, ended up swapping the modem for an 4G one. It seemed too bursty/congested to reliably do videoconf (the main reason we got an extra connection).

The thing that really left a bad taste with EE was the amount of my time they wasted, I can't ever do business with them again knowing they value peoples time that little.

> ”I could pull 100Mbit in the middle of the night but it would go down to 3Mbit in the middle of the day”

Are you sure you’re on 5G? Or maybe you’re quite far from a tower with a weak signal?

I’ve used both Vodafone and Three 5G in London and never see speeds that low with a good 5G signal. Certainly never below 100 Mbps, and rarely below 200, even at peak times. Rarely below 400 on Three, and often as high as 800!

On Vodafone even the 4G network is usually 30+ Mbps. But Three’s 4G is absolutely awful.

> Are you sure you’re on 5G? Or maybe you’re quite far from a tower with a weak signal?

I'm not using 5G, I was commenting on LTE in general competing with fixed line, but i'm using a Gigabit Cat20 modem with 2 MIMO antennas so it's not exactly slow and I have excellent signal close to many towers.

With this setup reception has never been an issue for me and UE throughput is good when there is capacity, as far as I can tell it's contention ratio, but I guess it's not necessarily that simple, admittedly I don't know how to tell if the contention is at the LTE part in spite of the signal and full carrier aggregation, or the backhaul (which I had previously assumed) and now you got me thinking.

> On Vodafone even the 4G network is usually 30+ Mbps. But Three’s 4G is absolutely awful.

Ya, this was my experience.

It seems that Three didn't invest much on 4G (hence the low speeds), but they fixed that with 5G. They now own a lot of spectrum (Three 160MHz, EE 120MHz, O2 100MHz, Vodafone 90MHz) and at least in central London you can see difference. On 5G, I often see 2x higher speeds on Three/Smarty than I do on Vodafone.

But again, 5G. Performance on Three's LTE is completely different. I believe they rank last on 4G speed tests here in London, behind EE, Vodafone, and O2.

I didn't have a positive opinion about them (used them for a year and then left, this was 5 years ago), but some people were getting interesting speeds with their 5G[0], so I ordered a Smarty SIM (30 day plan) just to test. While not as stable, it's certainly cheaper and faster than my FTTC connection (60Mbps/15Mbps for £30/month).

EE has always been very good in this area, even indoors - I was getting 150Mbps/15Mbps inside my flat, then moved to Vodafone which gives me similar speeds as my wired connection - but as you said, contracts are a problem. I'm not going do a 12 or 24 contract just to test 5G :-P

Anyway, if you have access to 5G equipment (a phone, 5G modem, etc) and have 5G coverage, test again. You may see completely different results compared to 4G.

---

[0] https://forum.kitz.co.uk/index.php/topic,23787.30.html ; https://forum.kitz.co.uk/index.php/topic,26761.0.html

I don't see how 5G will make Three faster if the bottleneck is backhaul, but maybe my assumption was wrong - I had excellent reception and maxed out the carrier aggregation on Three so it seems weird that it would be the LTE part that's the bottleneck.

If the problem in your area is the backhaul, then you're right, 5G won't help. However, at least in some areas, Three's problem isn't the backhaul. It's too many customers and not enough wireless capacity for 4G (and even 3G):

https://pedroc.co.uk/content/uk-commercial-mobile-spectrum

I've been testing Smarty (Three) and Vodafone in London. Almost every time Vodafone's LTE is faster and more stable than Three's LTE. On 5G it's different: not only Three has better coverage, but their speeds are often twice as fast. According to "Network Signal Guru" (needs root and Android, but it's an interesting app), Three uses 100Mhz of the N78 band while Vodafone only deploys 50Mhz. Not a surprise as they have more 3.5GHz spectrum than Vodafone, but shows the advantage they have on 5G.

By the way, I went to one of your early comments and you also mentioned EE. In the UK, networks share some infrastructure. EE and Three have MBNL ( https://mbnl.co.uk/ ) while Vodafone and O2 have CTIL ( https://en.wikipedia.org/wiki/CTIL ). If you live somewhere with shared towers/fibre and there are backhaul limitations, this may explain why you can't get good speeds on both EE and Three.

Cable internet is available too, I had gigabit DOCSIS from Virgin for the last few years.

You make it sound like the UK is in a terrible state for internet connectivity - it’s cheaper and better than many other countries by a long way, and those fibre providers which aren’t yet everywhere are insanely good where they are starting up. In the place we just moved out of, we could get 900/900 for around £20 per month!

> but realistically you need a ton more spectrum, which is why you have seen a spate of auctions lately.

There are huge amount of spectrums in the mmWave bands... instead of pulling FTTH to huge apartment buildings, just set a few fiber-connected towers, and tell customers to put the APs on windowsils pointing towards the tower.

Except, wires are better than wireless at transmitting information. Yes they require maintenance, but that's why we pay a service fee. Internet is a utility, it's high time we start treating it like one. When was the last time your water supplier started offering "premium ultra-filtered water"?

Put another way, finally more consumers will have more choices for home internet.

Yes, there are multiple home services providers. But they have carved - conspired? - to ensure that in many markets they don't actually compete. So where there could be more than one, there isn't.

Sounds like some political races ;)

Why is 5G being treated like a big public decision that we're supposed to have an opinion on, when in reality it's coordination between chipset manufacturers and cell tower operators that hardly involves us at all? It doesn't even seem like it's worth marketing to me, I'm not going to go out and buy a cell tower. Even selling me on 5G so I'll buy a new phone is unnecessary, the phone manufacturers could sell me on the improved performance directly.

The costs to carriers will be going up and by marketing they are subtly letting us know the cost of our service plan will be going up.

Don't think it's the whole story, but 3g, 4g and now 5g were used as big selling points for new phones and plans to consumers.

If people start buying things based on the name, they'll form opinions on it.

Most consumers purchase based on buzz words like 4g and 5g. They barely pay attention to bandwidth.

Because Chinese companies developed 5G before American companies did. American telecarriers not only didn't initially research the technology, they refused to even agree to a standard definition. IIRC, AT&T marketed their 4G network as 5G for a while. As a result, Chinese companies had an opportunity to enter western markets with superior, government funded technology. They started making sales, notably in Canada, IIRC, before Western pushback started. From the private sector, telecoms pushed on the government to ban the Chinese technology, to buy them time to play catch-up with China. Politically, the need to protect the inadequate technology (not) developed by the private market, from a government funded program doesn't fit with the current US ideological model, so a different excuse was found. The legitimate reason: "Telecommunications infrastructure is a national security priority, and Chinese technology in this arena, due to its ties to the CCP, is a national security threat" was used. Again, a legitimate reason. It even has the bonus that it implies the current western ideology that governments should not direct private research. But this sort of market protectionism makes 5G a public policy and national security interest matter. So in order to ensure there is public support for such a policy, one of the easiest things to do, is convince the bottom-information tiers of society with conspiratorial, xenophobic, nonsense. Which is where the 5G conspiracy theories come from. You will note that the 5G conspiracy theories can be generalized into the statement "5G is evil, and is used for mind control / cancer", which really, is a dumbed-down game of telephone (pun) away from the original source material: [5G is evil] <-- The CCP is bad / national security and we must reject their tainted technology; [It is used for mind control] <-- CCP = big brother, Chinese 5G = surveillance, national security; [Cancer] <-- more tangible than 'mind control'.

There still isn't a standard for what 5G is. Unlike 2G, 3G, and 4G LTE, which all have very specific technological definitions, 5G could be edge computing, MIMO, small cell, beamforming, and so forth. No one has actually created guidelines about what it is.

At present, it seems any kind of 5G service you can actually buy is at best just new frequency spectrums that the government auctioned off in recent years.

> There still isn't a standard for what 5G is. Unlike 2G, 3G, and 4G LTE, which all have very specific technological definitions

In common usage, these "generations" always match a change in the low-level protocol used by the radios. For 1G it was analog, for 2G it was GSM, for 3G it was W-CDMA, for 4G it was LTE, and for 5G it's NR. So distinguishing between 4G and 5G is simple: if you're using LTE, it's 4G; if you're using NR, it's 5G.

This is not really true – "5G" is pretty universally accepted to refer to 3GPP Release 15, defining 5G New Radio (https://en.wikipedia.org/wiki/5G_NR).

actually there is, ITU-R recommendation IMT2020 is 5G and only 3GPP Rel15 and following release satisfy this.

A lot of 5G conspiracy theories might also correlate with incumbent ISPs (especially cable ones) who can block new wired ISPs but they can't block mobile phones - so a lot of the conspiracy theories included things like "wifi connected by cable to internet is fine" :V

And 5G includes technologies that allow for much denser mesh, higher throughput, lower latency, and more clients per cell enjoying good service. So it could start supplanting wired broadband services.

I think expecting that most laypeople connect the Huawei ban to 5G as a technology is really strange. The conspiracy theories about 5G all relate it to wealthy Westerners like Bill Gates, not meddling by China.

It seems there are multiple consipiracies about it though. The Gov't is pushing that CCP will use it to infiltrate. This is more of a polictical paranoia consipiracy. That is quite different than the oridinary wacko crazy conspiracies being pushed by whomever is pushing them to say things like Bill Gates is going to do whatever for whatever reasons.

It is clear to me that the "5G conspiracy theories" were fabricated by American interests when Huawei were 5 years ahead of them and it was very possible that Canada, Australia, parts of Europe and a whole lot of Asia would install Huawei's 5G infrastructure.

It did backfire a bit when the American companies rolled out their half-baked 5G (which is absolutely horrendous and you are much better off disabling 5G on your phone for better connectivity and battery life)... but it was only a few burnt towers and not much else.

"Because Chinese companies developed 5G before American companies did" - just No.

5G as deployed today is a 3GPP (== driven by Europe) standard and yes, Chinese, American, Australian, Asian etc. companies contributed to system architecture and RAN (radio access network) which is a domain of the old-schoolers: Nokia, Ericsson, Qualcomm, Samsung and since a few years also Huawei-EUROPE (not Huawei China).

I remember there was a EU report that was suggesting Russia was spreading anti 5G conspiracy theories to slow 5G adoption so they could get ahead sooner. I searched for it a while ago but couldn't find it quickly.

The majority of the core 5G patents came from Qualcomm, Samsung, Ericsson, etc. Huawei filed a metric ton of land grab incremental patents on non essential details which is why they get less than 1% of 5G patent revenue.

https://www.statista.com/statistics/1237450/global-ip-revenu...

What you’re doing is running with this nationalist rhetoric on the Chinese internet that China “invented 5G”. Sorry, but 5G was not invented by Huawei. It’s fair to safe they rushed how a huge deployment of the technology, but mmWave is economically inefficient and most of the 5G benefit for China was increased subscriber density.

The reality is, the US doesn’t have the subscriber density problem China does, we are far more spread out and mmWave is much more useless to us, so one of the core bragging rights of Huawei is not relevant for our use case. The other 5G bands <6Ghz are much better for the US.

Age of social media. Everyone discusses topics they have zero control over to hell and back.

"Hey look, we're doing Democracy with thousands of people and everyone agrees/disagrees" - it really tricks the brain into thinking it's doing something.

Everyone's busy talking and believing they're doing something, we all feel good about it, there's no real need to actually do something.

Microcell deployment, regardless of whatever protocols they’re using, is part of the “5G” push. If carriers suddenly put one in your front yard [1] what are you going to do, be against technology and progress?

[1] https://abc13.com/houston-homes-seeing-5g-boxes-placed-witho...

Not to mention the ecological impacts of the density required.

Looks fine to me. Doesn't really stand out among similar utility boxes you find around for electric or gas. I'd love to have a 5g tower by my house and not need my repeater.

Carriers & phone makers need something to talk about for marketing. The iPhone 12 was unremarkable, so they had to lean on 5G as a reason to upgrade. 5G in my experience (latest iPhone 13 Pro on one of the largest US carriers) is broken/underwhelming, so I've turned it off.

I agree that there is a lot of bikeshedding regarding 5G and countless other topics (microservices vs monolith, electron vs native, etc). There are lots of topics that drive heavy user engagement while never answering an actual question.

~~Couriers~~ Carriers shouldn’t be in the business of selling phones at all, but that’s a different discussion entirely.

With net neutrality gone, once the dust settles they will also be in the B2B business of selling internet speeds to corps such as Netflix etc. There are already are so called fast/slow lanes

> broken/underwhelming

Really? It's been great for me. Tested it the other day and got 30 megs down.

I get flat out insane speeds on 5G. Must faster than any home internet I have ever even had the option to subscribe to. When it works.

When it doesn't work, my battery life goes to shit & I cannot check my email because no data seems to get through. I disabled it about 3 days after buying a 5G capable phone. I've never turned it back on.

I checked, 5G on and off. My battery drops 20% more when on 5G. iPhone 12.

So I wonder if all you smart people can do the math, how is that going affect power use and climate change when everyone has this new demand for power?

30mbps or 30MB/s? You can do 100mbps+ on LTE...

I get 300 MBPs download speeds on 5G - that might be a San Francisco thing?

Can't wait till its beamformed right into my eyebrain. If you are ever bored just look at the nearest tower and donate your wasted brain cycle towards SETI while you stand motionless drooling at the sky.

MB/s

Will it still be 30MB/s when 100,000+ people in your 5G cell are using it though? I don't understand the tech. 100k people in San Fran pulling 30MB/s would require 3 TB/s of spectrum bandwidth. Isn't that a lot, or am I just way out of the loop and still thinking in 1990's units. :)

Spectrum is weird. Technically WiFi is also spectrum and it supports every household in SF with way more than 30 MB/s each.

So it’s really just a matter of how many antennas and how “far” (including walls) can the frequency travel.

Oh that's a great example. The 2.4 GHz IMS band used by Wi-Fi literally serves hundreds of millions, but only have a radius of about a house. I heard 5G cells have smaller diameters, so that makes way more sense. Thanks!

The reason why mobile internet in rural areas is bad is because carriers are too cheap to put cell towers every few kilometres.

They should really be forced to in the same vein as how telephone companies are forced to run service anywhere.

Broadband should be declared an essential service, like power and water.

iPhone 12 was huge upgrade from 11 even without 5G. It adopts OLED. slim bezel, became lightweight. iPhone 11 was tiny upgrade from XR but people think tend to think it was big. It was just a recovery from failure of product naming of XR/Xs.

"Even selling me on 5G so I'll buy a new phone is unnecessary, the phone manufacturers could sell me on the improved performance directly."

If 5G makes it cheaper for the carrier, they could make plans cheaper for folks whose handsets support 5G, i.e. penalize people who don't upgrade.

Wait, did you just suggest a carrier would make something cheaper, voluntarily?

I seriously doubt that. Even if it does lower their operating costs, why would they lower their potential profit padding?

Competition.

Why do Visible and Mint offer cheaper plans than the Verizon and AT&T brands?

To attract customers from other carriers.

Do either of Visible, Mint, etc install their own infrastructure or buy bandwidth from ATT and/or Verizon?

Also, their services are spotty at best compared to other providers. There's a reason for their prices being lower similar to how Spirit Airlines is cheaper than other carriers.

Mint uses T-mobile as its underlying carrier.

"their services are spotty at best compared to other providers"

How so?

AIUI the only time Mint customers would have worse service than T-Mobile customers is if there's network congestion. e.g. at a crowded sporting event.

I have friends with Cricket (not sure which carrier they attach). In the city that they spend the majority of their time, the service is fine. We went on a road trip where once we left the major city's coverage, their signal was no longer very good until reaching another major city. My phone with one of the main companies where they own their network never lost signal.

That's the anecdotal consensus I've heard from all of these smaller vendors.

Visible is part or Verizon, i.e. it's just another Verizon brand. I'm guessing they save a lot of money on customer service, as only relatively savvy customers will sign up for a mobile plan online, without talking with someone or walking into a store.

The top line metric that the financial markets care about for the major postpaid carriers is ARPU (average revenue per user). As soon as this number goes down, so does the company’s market cap. The C suite really cares about making sure this number continues to go up.

The easiest way to improve ARPU is to raise all existing customers' prices to current ARPU, and hope they either accept it or churn. Either way ARPU goes up, whilst revenue and profits go down.

I'm pretty sure the carriers want both ARPU and profits to increase.

If a carrier CEO has two options, with the following estimate outcomes:

A) Profit +10%. ARPU -1%.

B) Profit +5%. ARPU +5%.

Do you think they might choose A?

There is little doubt that both of those companies are lowering prices because of the data they collect and sell.

> Wait, did you just suggest a carrier would make something cheaper, voluntarily?

Comparatively cheaper. You can do that by raising the price on the alternatives.

Adjusted for inflation, mobile internet has been decreasing in cost - I pay the same price I did a decade ago and have much better coverage and much better speeds.

Carriers are marketing 5G as a public facing feature so they can charge more for it.

Spectrum allocation in licensed bands is a public decision. The public literally decides what use those chunks of spectrum should be put to. Only the free bands (like 2.4GHz) are a pure matter of cooperation among users and hardware manufacturers.

I thought it was decided by an auction?

Who gets specific assignments is often done by auction, but how the overall spectrum is allocated is by the FCC. In theory they're responsible to the public.

That implies 1) the FCC listens to the public and 2) the public understands the technology and 3) the public will vote for what's good for themselves in the networks domain. Arguably 2 and 3 are false, and 1 we know to be false depending on the admin (see: net neutrality)

I'm having deja vu, didn't Ajit Pai ignore all that public input from the web forms (blaming it on prank submissions)?

I don't think it implies any of those things. If something is regulated by the FCC, it's regulated by a public agency, even if that agency disregards public opinion or public interests. If an anti-5G sentiment caught fire, the FCC could be directed by an administration or Congress to move in a different direction.

Anti-5G sentiment shares a lot of DNA and membership with antivax sentiment which seemed like a vocal fringe minority 20 years ago, but may end up being a majority of voters in the next election cycle. It's not completely out of the realm of possibility that you could see major anti-5G political candidates and/or notable incidents of anti-5G violence and terrorism.

The irony is that the main case for 5G is reduced costs for operators.

4G blew the top off what people thought the Shannon Limit was but at the expense of extreme and expensive coordination between base stations. (If your digital front-ends are sampling at a rate of N Hz and a bit depth of d you need to bring all that data to one place!)

5G gets better spectral efficiency (serve more customers with expensive spectrum) with a simpler coordination model. (lower capital cost)

Of course since 5G is the new shiny carriers want to charge you more for it but they'd save money if they got you off 4G and onto 5G.

It most certainly did not do that, neither has 5G for that matter.

The Shannon's Law refers to https://en.wikipedia.org/wiki/Shannon%E2%80%93Hartley_theore... to see the actual formula

the basic principle is that the Capacity C of a comm channel is proportional to the amount of spectrum you have B and the Signal to Interference and Noise ratio (SINR).

So basically, if you want more throughput, you need either more spectrum or more cell sites in order to improve the SINR, you cannot simply continue to crank up the power because then you are increasing the co-channel interference (interfering with other terminals on the wireless network). This is why Spectrum is so expensive, the alternative is to build more cell sites, which is expensive.

4G LTE did come a lot closer by using OFDMA but they most certainly did not exceed the limit. 5G using mm-wave also did not, it uses phased array antennas to implement beam forming which created space separation, that is, each user gets their own beam and do not have to share it with others, but within that beam, they still are confined to the law C ~ B * SINR

The operators did add a ton of spectrum through various auctions and that plus the new tech is what is driving the tputs up.

You need more SNR (power), more BW (spectrum) , or more 'channels'. 5G solves this using the last dimension, channels. Beamforming allows the same spectrum to be reused by allowing for spacial separation, i.e. use the same spectrum 5 times by sending 5 beams to 5 physically sperare locations. MIMO is layered on top of this ans uses multipath effects to offer more bandwidth to a single user by taking advantage of the fact each path is its own channel. No violation of the Shannon limit for sure, but very smart and computationally intensive techniques to skirt around the issues of the single isotropic or sector antenna of old.

3G and prior "cellular" systems were inefficient because they can't reuse the same frequency in adjacent cells. See the diagram here

https://www.geeksforgeeks.org/frequency-reuse/

Here you have 7 different frequencies used in different groups of cells.

In 4G all frequencies can be used in all the cells. If you are between two or three cells you are probably receiving a signal from (and being received by) multiple cells and they are sharing the RF baseband to make it possible. That plus a big bag of tricks let 4G achieve radically better spectral efficiency through "spatial diversity"

> Here you have 7 different frequencies used in different groups of cells.

You should only need 4 different frequencies: https://en.wikipedia.org/wiki/Four_color_theorem

For hexagons you can get by with 3, https://en.wikipedia.org/wiki/Hexagonal_tiling#Uniform_color...

You'd think so but practically the signal bleeds across to second-adjacent cells and a real channel plan needs more than that.

Look at how the 3 WiFi channels in 2.4GHz land is nowhere near sufficient for good spatial diversity and it is even a struggle with the 5GHz channels that aren't gated to avoid interference with radar.

Indeed. In fact, the link I responded to shows a pattern matching the 7 color 3-uniform distribution I linked to

That only works if you have exact edges and that each BTW uses only one base channel (iirc, they mix and match as much as possible)

Yes, each cell interferes with the next, that is call co-channel interference. That is a major factor that limits the capacity of a cellular network. There are a number of approaches used to manage that, carriers spend a lot of money and employ many engineers designing around that.

SON (Self Optimizing Network) is the latest tech that is being deployed on 4G and 5G to help combat that.

BTW, this is one of the reasons you can get much higher tputs with mm-wave because the beams do not propagate very far which makes keeping co-chan down, that's the good news, the bad news is it doesn't propagate well, which means you need way more nodes.

It's not a hard stop between cells. You don't have radio signals and then suddenly nothing because of reflections, absorbtions and the very nature of log scale of signal distance drop off..

4G also reuses the frequencies in the exact same way, the primary difference between 3G and 4G is that 3G used CDMA and did not have the higher modulation desities, (capped at QPSK) while 4G LTE went up to 64QAM, so more bits per symbol.

4G did not use spatial separation, that requires beamforming, that is most certainly not deployed in 4G, 4G LTE uses MIMO. Spatial separation is deployed in 5G but only at the mm-wave bands due to the 1/2 wavelength separation required between antenna elements, hence, impractical with longer wavelengths.

3G (UMTS) was reuse 1, not reuse 7 !

This is not the only mistake from this website you mention : e.g. in the page "difference-between-gsm-and-cdma" I read "GSM Data speed rate: 42Mbps in HSPA" (but HSPA is 3G and an evolution of UMTS and is using WCDMA, compared to GSM which is 2G and is using TDMA). I guess when they mention "CDMA" or "3G" (which is are generic terms that also apply to UMTS) they mean "CDMA from 3GPP2", which is not used in Europe. I didn't read everything but from the little I have seen I don't find this website very reliable...

I am out of my element here but I was under the impression that the capacity of an arbitrary wireless network was still an unsolved problem. (Though I presume what you are referring to is a case that is solvable)

https://en.wikipedia.org/wiki/List_of_unsolved_problems_in_i...

hmmm, i'm paying exactly the same rate I have been paying for years and even took them up on a second business line for free. Your theory isn't panning out for me. Also my 5g experience with t-mobile is that it's between 3-8x as fast as I was getting before on LTE in the areas that I frequent. So I'm not complaining. I mean I WANT my gigabit 5G but I'm okay right now at this pricepoint.

Ugh, that's just super opinionated and short sighted. I have countrywide 5G here, and the difference in latency on voice or video calls is super notable and nice.

Obviously it doesn't make sense to run on 5G permanently because energy consumption is still comparably higher, just like it was the case with 4G in its beginning, but the automatic switching does a somewhat good job at it.

5G also has a ton of other benefits such as being able to support a multitude more clients and being able to prioritize traffic (like for emergency services) and handle slicing/QoS much better.

Not to mention stuff like e.g. active beamforming capabilities reducing power consumption and lowering emission levels for the general environment SIGNIFICANTLY.

All of the amazon internals and obvious marketing/hype around 5G aside: Author is a good example of naysaying - Saying no to obviously better technology because it doesnt fit their reality and/or knowledge.

What kind of latency are you seeing? I see 4G latency around 30±10ms so I could believe that's an area for improvement but also one which wouldn't be transformative for many applications.

From my perspective, the big change we need in the U.S. is getting away from tiny data plans — most of the interesting 5G applications also use a ton of data and when 1GB is concerned a princely amount it's hard to care much about video calls.

I get ±40-60ms on 4G legacy and ±5-10ms on 5G. Which no matter what people try to claim here is a great and very notable improvement in latency on voice calls.

Another benefit I have noticed is that latency seems to be much more stable in all kinds of conditions - stationary, in trains, in the car. I believe this is one of the key factors in what makes the experience so much nicer.

I have no idea about US cell plans, here in NL it's included in afaik all new contracts. I have unlimited data + 5G for 40€/month

That does sound like a nice improvement. We’re held back by the pricing model so most people I know have treated 5G as something they’ll get eventually but won’t really change what they do.

Reminds me of the folks on my neighborhood's Nextdoor that claim that their Hughes Net connection (1s+ ping, tiny data caps) is just as good as Wave cable (xxms ping, relatively huge data caps, lower price).

being able to prioritize traffic for emergency services etc.

Emergency priority been in every mobile system since AMPS.

5G has a few use cases. In very high user density areas, such as stadiums, it's possible to have a huge number of short range connections in the 24-40GHz band. This allows people at the game to watch the game on their phones. Many major stadiums installed this back in 2020.

If you want huge bandwidth, you're going to need near line of sight to the base station. The high band won't go through much. Stadiums and entertainment venues are the ideal case, because they're designed to let a large number of people have line of sight to something. So they were done first.

Worst case is a subway or complex building interior. Samsung has little brick-sized beam-forming units.[1] It takes a base station every 150-200m or so to provide good coverage in the high bands. AT&T is integrating them into street lights.[2] Not clear how they backhaul.

[1] https://www.samsung.com/global/business/networks/products/ra...

[2] https://5ginsider.com/carriers/att-rolling-out-5g-powered-st...

> This allows people at the game to watch the game on their phones.

I'm not a sports person, but I'm having a difficult time understanding why someone would want to do this. Even with the very best e2e latency, there's probably going to be one or two seconds of latency between real life and what's on your phone. Sports is very much a real time experience, and that sounds lame.

I am a baseball fan, so maybe other sports are different, but from my perspective: When I go to the stadium, I have a single vantage point. Oftentimes I'll have a low resolution stream going of the game where a few-second delay is optimal because I'll get to see a replay of something that happened out of my view or that I just missed. Plus, for all but the biggest plays, a replay of a close call isn't usually shown on the scoreboard.

This is not conclusively true. 5G also uses sub-1GHz bands for long-range connections and these will also support a multitude more connections per tower and provide lower latency.

Please refer to [1] to understand how much more advanced traffic prioritization is implemented in this protocol.

[1] https://nms.kcl.ac.uk/toktam.mahmoodi/files/EW16.pdf

> Obviously it doesn't make sense to run on 5G permanently

I've never used 5G (it isn't available in my country) so this isn't obvious at all. Care to explain that doesn't make sense?

If your phone is just sitting on your desk with its screen off there is little to zero benefit to having it run on 5G, I mean. As of right now, 5G is still somewhat more energy consuming than 4G, just like 4G consumed more energy in its early stages.

The "Why Not?" section here starts with a disclaimer that the author is personally affected but doesn't actually say how. The points raised boil down to "I have enough speed so I don't see the point". Ok, but that's not a reason why not.

Remember bandwidth isn't just about how much an individual can get but how many people can be serviced within a given area either in total or without subdividing into cells, which then require somewhere to put another tower.

The author didn't raise this but I've seen others who have brought up the nebulous "radiation" argument against 5G. Here's a good litmus test: if someone can't tell you, at a minimum, what "ionizing radiation" is then you can safely ignore everything they say about "radiation".

Here's another: if they can't describe the "radiation" in terms of the radiation exposure from eating a banana, you can also safely ignore them.

While you're ignoring everything they say about radiation, you can also suggest they watch "All you need to know to understand 5G" by Sabine Hossenfelder.

https://www.youtube.com/watch?v=FBsP-bmDLOo

> The author didn't raise this but I've seen others who have brought up the nebulous "radiation" argument against 5G. Here's a good litmus test: if someone can't tell you, at a minimum, what "ionizing radiation" is then you can safely ignore everything they say about "radiation".

I often use that test myself, or some variation thereof, However, it doesn't always work. We often regurgitate knowledge from people that couldn't recite precise definitions. As a matter of fact, I think most of the people around me, as well as myself and, I think, a good part of commenters here, wouldn't be able to properly describe ionizing radiation, but yet wouldn't refrain from commenting on related matters. On top of that, I met quite educated and convincing persons bring up (non-ionizing) radiation arguments about 5G, that wouldn't have any problem telling you what ionizing radiation is.

I've got two friends that I know are more-or-less 5G skeptics, and heard them sharing how frustrated they are that they cannot even give their opinions on these subjects without being ridiculed, so much that they even lie when asked about it, depending on who's aking them, and then watch others spread vaguely similar conspiracy theories.

That was quite terrifying to me (at the time).

> Here's another: if they can't describe the "radiation" in terms of the radiation exposure from eating a banana, you can also safely ignore them.

As far as I know, most people that are (seriously) bringing up radiation arguments about 5G are worried of non-ionizing radiation.

> I've got two friends that I know are more-or-less 5G skeptics, and heard them sharing how frustrated they are that they cannot even give their opinions on these subjects without being ridiculed

There's not really much to be skeptical about with respect to the radio aspect. You're illuminated by far more watts of radiation from non-5G sources than you are 5G sources, more so if you're outdoors. Not only is the 5G radiation non-ionizing but it doesn't penetrate for shit through water (a major component of humans). Unless you put your face against a tower's antenna there's nothing about a 5G radio worth worrying about.

As for the network part, it's a high bandwidth but low latency connection rivaling wired internet but wireless. It's no less useful than high speed wired Internet. For some applications, mobile but bandwidth or latency sensitive, it's infinitely better than wired.

Microwaves aren’t ionizing radiation, but you still probably don’t want too many watts to the brain.

That said, I have no opinion on 5G safety, I don’t know enough to have one. Do you know what the wattage of a microcell that might sit on a suburban power pole might be?

And don't forget the inverse square law. You will be irradiated much more by a cell phone pressed against your skin than a microcell on top of a pole, just because the distance is many orders of magnitude more.

The arguments that I like to use are:

- The Sun puts out vastly, and I mean vastly more radiation by every metric, much of which is the ionising, cancer-causing type. The same people frothing at the mouth about 5G will spend hours suntanning.

- Mobile phones are deadly, and cause a surprising number of hospitalisations annually... from car crashes due to texting-and-driving. If you care about mobile phone safety and your health, why aren't you focused on that?

- If you can't feel it burning you, then it's not putting out enough thermal power to burn you... by definition. Analogies with microwave ovens are irrelevant when the oven would give you a severe and immediately painful burn if you put your hand in there.

- The main difference between 4G and 5G is the protocol, not the type of radiation. People think 5G is some unique thing like "X-Rays but different" or something. It's literally a radio. A radio! People live near radio towers putting out hundreds of kilowatts and are fine.

- A key benefit of more efficient protocols is that they need lower power to send the same data. So any argument for going back to 3G or 4G to avoid 5G is an argument saying that more radio power is... better somehow?

All of this assumes we’ve reached some plateau for how the Internet will always be used. Historically that’s never been the case — as more bandwidth becomes available, new applications pop up to fulfill the need. Perhaps it feels like marketing fluff now, but generally reducing latency and increasing bandwidth has always historically lead to new kinds of use cases that become daily drivers. What would be newsworthy is if that didn’t happen.

> Historically that’s never been the case

We overestimate how old history is here. There is a case to be made about how we've enjoyed exponential growth in cosumer technology over the last few decades but that could slow down on a few fronts. For ex- display resolution has reached "good enough" fidelity for a while.

I certainly hope you're right and we find cool, novel use cases but I wouldn't be certain. I personally have not thought about bandwidth for a few years now. Meanwhile I remember the speed bumps being exciting earlier. Diminishing utility is real.

I still think about throughput and bandwidth from time to time. In my case, I am mobile enough to encounter the fast but high latency case and fast, but only in little bursts, too few to gloss over, cases. There is room for more here, and it might lead to something new, that scales.

That said, I do agree with you.

In my view, what the carriers do matters more.

They are still wanting to gatekeep to a much higher degree than they currently are.

Nothing pays like creating problems and then selling solutions does. Massive consolidation opens the door for the threat of artificial value to walk right through and into our wallets.

Take the truisms in my other comment: https://news.ycombinator.com/item?id=30822400

That quality vs choice problem can be made to go away for what could be called a nominal charge each month. Think all the ugly FastPass was for Disney, applied to our mobile experiences. It's very ripe fruit, sadly.

> display resolution has reached "good enough" fidelity

Ehhh it was only a couple of years ago that 4K became affordable at 120hz. I would absolutely buy a 5K display if one with decent input lag and no DP compression existed.

You can go 100 years and see this, from telephone lines that were only local, to digital switched and dedicated lines, to satellite coverage, and all of the applications as we’ve gone from wired dial up internet to being able to FaceTime in the middle of the street. The same thing has happened with TV and it’s bandwidth increases on airwaves from just a few channels, black and white to color, cable, satellite, and now streaming online in 4K, soon 6K and even 8K… let alone in VR. The history isn’t so new.

I would not take your bet the world is good enough and innovation is dead. That’s been a wrong bet for thousands of years. Every time a new discovery or technological advancement is made, new use cases come that then fuel the next one… from the Bronze Age to stainless steel and fiberglass, medicinal herbs to antibiotics and mRNA vaccines, writing systems to tablets, paper to the printing press and now this comment on HN.

> has reached "good enough"

That was said about fullhd and even hd-ready. Fast forward to now and, no, it is not good enough anymore. Same about TN displays, same about 60Hz, same about 300 nits of brightness.

Or as another closer example: awful upload rates were good enough when one only needed consumption, but now with home office work upload is also important

The carriers themselves have an impact on this beyond physical layer improvements.

A big one is data cost rates and caps.

I've heard more than once, "5G means I will zip right through my data and will cost me more."

LTE actually works great for me, and I do video calls and such over LTE fairly regularly. Now I do pay for that, and I'm compensated. That's what the carriers would love to see continue, but that also limits mass adoption too. Many people don't like extra charges appearing on their phone account. I don't, but I'm in a position to deal with them too. Annoyance.

Plenty of people really feel that extra $20, $50, more depending on what it is they did, like drive too close to Canada with automatic data roaming turned on. That can be painful.

Years ago, early 90's era, there was a magazine. Boardwatch, I believe it was. There were some great voices in that one, and all these topics we discuss today are nothing new, and most of us here know that.

There are some constants:

As mentioned, applications fill resources. Could be RAM, storage, throughput, bandwidth. Part of that is offering choice, the next one coming up.

*In broadcast / streaming, where there is a defined bitrate possible, choice will win out over quality. Here's a little thought experiment that suggests why that is generally the case:

Say you've got two streams or broadcast channels to use. One is boring, but exemplary quality, the other is poor quality, but compelling as it gets. Which one do you use? Which one do you believe most people will use?

That leads to, "Content is king."

One of the topics Boardwatch covered regularly was diversity in ISPs. As consolidation happened, another truism did as well, and that is basic economic trope at this point:

More choice = more competition = buyer gets highest value for the dollar.

Less choice = less competition = buyer gets lowest value for the dollar.

The fight to prevent lock-in is therefore eternal. Carriers of all types will seek to limit choice however they can.

Higher cost of change = lower choice.

This is also driving the one I mentioned above. Content being king means making more content options available almost always trumps quality.

I do have a small nit to pick:

To me, the word "bandwidth" isn't quite right. Yes, it speaks to the maximum information possible, but throughput is the more accurate word, in my view, when it comes to the impact moving more bits per second at lower latencies has on society overall. Another quick thought exercise:

Say you have a choice of a very fast connection, but latency is all over the place, and or there are random slowdowns vs. one that isn't as fast, but is super consistent. Which do you use? Which one do you believe other people will use? I prefer the latter most of the time given the constancy metrics match up with my use cases.

My point here is the carriers have a huge impact today! We've consolidated down to a point where many truisms about the Internet are not so true in the wireless realm. Back in the 90's, when most of us were on wires most of the time, all that discussion about carriers, ISPs (when they were two different things often enough to make that distinction), and the up and coming "cloud" computing being equated to how things were in the late 60's and 70's, has played out fairly accurately.

Kudos to the people with vision back then. Too bad it didn't have a bigger impact on public policy overall.

> Historically that’s never been the case — as more bandwidth becomes available, new applications pop up to fulfill the need.

You're writing this on HN of all things, where your comment would take up no more bandwidth today than in the glory days of dialup, and only a bit more than serving it via a BBS or UUCP.

Loading this today is more/less instant. Even on dialup it would have been seconds.

And HN-style content does not represent the majority of bandwidth usage, or the new applications that make use of higher bandwidth. TikTok simply couldn’t have existed in the dial up days, a company now worth billions. That has happened consistently with communications technology over the past century. Why would we think it should stop now?

The author went the entire article without actually defining what 5G is. If he had, it would have made some of the discussion points more obviously moot.

4th generation has limits on total devices, power utilization, bandwidth, latency and coverage that 5th generation improves upon.

Carriers are the ones who want to move to 5G - they are clubbing phone manufacturers to do so, and then both are marketing it so people buy into the new infrastructure.

You'll see other groups (smart cities, autonomous vehicles and private IoT deployments) with technical need. In the US, we saw T-Mobile push for lower frequency usage at the start to improve their rural coverage, while their Sprint side pushed for more of the bandwidth-improving urban coverage. The article has several quotes from people who are using 5G for cheaper last mile for home internet - this could be a mobile hotspot or an antenna installation. This will provide more competition for internet service in areas that have had effective monopolies on broadband.

The towers and frequencies are being switched over to 5G, with emulation for handling older technologies. I would expect as the networks start to go past the capabilities of LTE, there will be congestion controls which will give LTE customers less bandwidth or lower quality of service at times.

> The author went the entire article without actually defining what 5G is

Iam pretty sure they actually dont know.

> Is 5G a cheaper or better way to do that? I don’t know, but it doesn’t sound crazy.

> The phone calls what it sees “LTE+” (I don’t claim to understand what that means)

This is a guy who asks their twitter audience and wonders why there is no explicit answer.

It seems like hubris. I see it a lot in very senior engineers. Basically, the belief that "I have all the information in my head to answer any tech question". The blog post would have been much better if he spent 30 minutes less on twitter, and 30 minutes more on familiarizing himself with the problem space.

Announcing you're a skeptic about something, and then having your analysis filled with knowledge gaps isn't a great look.

Reminds me of...

"Who needs faster CPU cycles?"

and

"Who needs that many pixels?"

I'd suggest the question should be something like "What does the lifting of those constraints enable?" or "What businesses will this make possible?"

That's just the same question phrased differently. I'd suggest that what we need are answers.

It wasn't intended to be. The point is that in those circumstances when we were expanding the resources we had been accustomed to there were many people who said we'd never use them. Disk space is another. However, the commoditization of a resource or the elimination of its scarcity invites us to use the resource in a way that would have seemed wasteful previously and perhaps may have blocked certain use cases before. The games industry has been loved and subsidized for driving consumption of resources up to drive new hardware adoption.

Another example of this is disk space where it's reduction in cost has driven changes in the industry. Think of NoSql and other shifts due to the reduction in cost of storage.

Diminishing marginal utility is certainly a thing, and I would counter argue that the longer it takes for a technology to be transformative the less likely it is actually needed by society. At a certain point in time the answers to both of your questions are going to be "no one". Do you need a 32k resolution screen? Do you need a 8.4GHz processor? Maybe, maybe not. If you keep bumping the numbers up you will eventually see an imperceptible difference in performance that won't justify the costs to build it from the oems.

The only real use case I've seen for 5G has been people live streaming themselves from a cell phone in major cities. That didn't seem possible before, but so far it's a small niche not something transformative. The only other use case I could think of was streaming live sports from people's cameras and making a digital television network out of that. However, I think you would likely get kicked out of every stadium in the US if it caught on as you would be undercutting the major source of revenue for most sports leagues.

32K or higher resolution screens might allow for wall displays and interactions with subsets of the screen. What does that enable? I imagine things I cannot yet imagine. A remote dance party may feel less odd and people you are remotely meeting may seem more "there" if you can see them moving in the space. Intensive resolution might increase the realism of virtual stimuli. I don't know... there's no magic in my brain.

The point was that many of us were previously unable to see what use an expansion of capacity would allow. This it seems reasonable enough to think we might have a crisis of imagination again.

I moved house recently and in a reasonably large suburban area was forced to rely exclusively on tethering with 4g/5g on my iPhone until the fiber could get moved to the new place.

It was a miserable experience. Video calls (wfh) were impossible- laggy and cut out all the time. Several software updates happened that week - at about 1gb average for almost a dozen devices - personal and work MacBooks, and phones for a full family with kids, I burned through my tethering allotment for the month within a week.

Latency with 5g is better than LTE but the best I’ve seen is in the neighborhood of 30ms. I can easily get sub 5ms on fiber. It does make a difference when doing rdp sessions for example.

I get that fixed wireless may work for some but for us it’s a non starter after being used to fast reliable fiber for almost 15 years now.

Another anecdote — at my house, on Verizon Wireless:

LTE: Consistently 5-20Mbps down. In the first couple of months after LTE went online here (5ish? years ago), I could get more like 90-100. It dropped into the 20-ish range and never went back up.

5G: I have UltraWideband coverage at my house, and I can get 200-500Mbps download speeds.

Will 5G just drop down over time (as usage goes up) like LTE did? I dunno, maybe. But, for now at least, 5G(UW) has been much, much faster for me.

Well, that's almost entirely due to Verizon's C-Band being pretty empty right now (5G phones are only a couple years old) and it being 2-5x the frequency and bandwidth of the bands they put 4g on. If they had put 4g on C-band, you'd probably be seeing the exact same performance.

So, I guess what I'm saying is, if by "5g", you mean "the giant new swaths of high-frequency spectrum that carriers are building out", then 5g is amazing. Otherwise, it's nothing to even bother noticing, as a consumer.

Definitely the biggest boon of 5g is the high frequency spectrum.

I guess that’s kinda my point — for me, “giant new swaths of spectrum” is my 5G experience. (Just a part of how muddy the waters of this topic are.)

Yeah, for me too. I'm excited about mobile again for the first time since LTE got turned on.

Agreed. We have a limited backend that appears to be expanding at a snails pace.

There's a stretch of driving I've regularly done which goes through an unpopulated stretch of the US. I've watched the tech tick up from 2g, 3g, 4g, LTE, and now 5g. Yet the download speed remains at sub 1Mbps speeds.

My assumption is they are still running the same copper/fiber that was buried to the tower 20 years ago while updating the receivers.

If they are running 20 year old copper that will slow things down, and your only solution is to replace it.

20 year old fiber? Upgrade the optics on both ends for a couple grand and you are up to date again.

Hi, another 5G UW user here, I am getting pretty consistently 1.8Gbps to 2.1Gbps on mine. But I played around with the CPE placement quite a bit. I don't know if you are able to try that. UW will drop you down to lower speeds or a lower band if the mm-wave signal is not good. mm-wave is really touchy dues to the small beam size.

Sufficient 5G saturation will likely result in induced demand. Skepticism is similar to being skeptical that anybody would ever need more than 64KB of memory. Right now, 5G saturation isn’t large enough for new applications to be developed that assume 5G, but once 5G is everywhere, we’ll look back on our pathetic tens of Megabits that we pay $40-80/mo as the Stone Age.

Things that more bandwidth provides: wireless home internet, cloud gaming, 4-8K streaming on multiple devices. Things that 5G provides: more efficient wireless connections which should, in theory, eventually get passed on to consumers.

Many of the 5G applications do exist currently, just in their early stages. Just like 5G penetration.

5G is just a larger wifi basically, you trade high speed for short distance, it is useful for hot area(populated places, campus, factories, stadiums, etc), but it does not make much sense for wide spread locations, for those 4G is enough and much much cheaper with a much cooler radios(and in much less quantities) on cellar towers.

IMO, 4G is already an optimized tech considering its price, distance, speed,etc. in the future, it's more likely 4G will still be dominant while 5G can used in those hot zones where 4G speed is not enough(again, the crowed areas as mentioned above).

For me 5G has been absolutely wonderful. It made my connection at home about 2x faster than whatever we have at the city office. And about 50x faster than what I had at home before. Also much more stable latencies.

You must have some pretty crappy wired ISPs

Perhaps. There was some indication that the wiring in our house caused some trouble for our previous VDSL.

But even if it had worked at advertised speeds, 5G would have given me roughly double the speed. As it was, 5G gave us about 50x better and stable connection. Only fiber directly to our house could top that, but unfortunately that's not available here yet. Plus costly.

> For me 5G has been absolutely wonderful. It made my connection at home about 2x faster than whatever we have at the city office. And about 50x faster than what I had at home before. Also much more stable latencies.

But why are you using 5G at home in the first place? Where are you that 5G actually outperforms your wired connection (accessed through WiFi)?

It's cheaper than fiber internet here. Costs less for the same bandwidth, without a contract and without data caps - unlike the wired option. Assuming the latency is usable, unlike 4G, it's starting to be a very competitive option. If you're stuck with 12mbps copper like some areas, it's a very obvious choice.

Suburbs.

I don't understand the doubt. Why does everyone expect that 5G wouldn't work for this case? It's just a way to connect the last kilometer to a fiber network without expensive (and in my case, unavailable) cabling work.

I think the issue is that 5G was initially rolled out with hand-waving and marketing. The initial 5G rollouts in the US have been on low-band spectrum (below 1GHz) with companies using 5-15MHz channels (or in the case of Verizon, not even giving dedicated channels to 5G and using inefficient dynamic spectrum sharing with LTE) and most of the bandwidth that users get is actually coming from the LTE channels on the network via carrier aggregation. So, out of the 40-50MHz of downlink capacity your phone is using, most of it isn't 5G and low-band 5G's efficiency improvements over LTE are a bit minimal (probably 20-30%).

Mid-band 5G is another story. With better MIMO and such, the efficiency improvements seem to be more like 50-80%. Not only that, a lot of new spectrum is being added to networks and it's TDD (time-division duplexing) rather than FDD (frequency-division duplexing). T-Mobile and Verizon have launched their 2.5GHz and 3.5GHz networks with 60MHz of spectrum and people are usually seeing speeds around 300-400Mbps which is significantly faster than LTE (and significantly faster than the average home broadband connection). A lot more mid-band spectrum will be coming over the next couple years.

Likewise, initial rollouts of 5G were in non-standalone mode. That means that LTE is running the show and the 5G is providing its bandwidth. That also means that ping times are driven by the ping times of the LTE network. 5G standalone networks drive down ping times a lot. Network and handset support is a bit iffy right now, but it'll come along.

I think it's also easy to forget how LTE was a minimal upgrade over HSPA+ when it initially came out. Why not just use dual-carrier HSPA+? Even in mid 2013 (2.5 years after Verizon's LTE launch), they were averaging 10.25Mbps while T-Mobile's HSPA was 7.66Mbps (via PC Magazine). Only a third faster? Why bother, right? Of course, that hides the fact that LTE drove ping times down a lot from 3G networks which meant a much better experience for users. It also hides the fact that LTE had a lot of room to improve over the years and now we see LTE speeds 4x higher than 2013. When you're comparing a newly launched product to the last generation that you've spent years working on, the new product can often seem like a minimal upgrade (even if it has a lot more runway).

T-Mobile's mid-band network is the most deployed in the US and their average speed has risen to 91Mbps (via Ookla/Spedtest.net, as of 3-6 months ago) - not their average 5G speed, their average speed. That's with less than half their network usage going over 5G and even within that 5G usage, a lot of it is probably over the low-band 5G network (which is more like LTE++). Their average 5G speed is 187Mbps and that's averaging in both the low-band LTE network and mid-band LTE network (again via Ookla, as of 3-6 months ago). I think that going from 40Mbps to 190Mbps is a pretty big difference.

Not only that, it's a better real world experience. Ookla found that 13-20% of the time (depending on carrier), LTE was below 5Mbps down and 1Mbps up for people. T-Mobile's 5G speeds were at least 25Mbps down and 5Mbps up 82% of the time. So with LTE, you're going to get at least 5Mbps 80% of the time and with 5G, you're going to get at least 25Mbps 80% of the time. Plus, 5Mbps up means a nice video call compared to 1Mbps. 5G is offering better stuff in the real-world and this will get better as the networks mature - carriers are still upgrading towers.

Now, if you're on AT&T, they have almost no mid-band 5G. If you're on Verizon, they've just launched a decent amount of mid-band 5G (30% of Americans covered, although people are reporting that their maps are hugely over-estimated), but it's limited compared to T-Mobile and only available on their premium plans. It might not seem like a big deal if you aren't seeing mid-band 5G.

I think 5G is also starting to put pressure on home internet providers. T-Mobile gained more home broadband customers in Q4 2021 than anyone (wired or wireless). A lot of rural customers are getting real broadband for the first time and upgrading from things like HugesNet with restrictive data allotments. Starlink "isn't a big deal" if you already have good broadband. If you live in an area without it, Starlink gives you a modern internet connection. Likewise, wireless home internet will offer new options to many rural customers as well as put competitive pressure on broadband monopolies.

I think part of Tim Bray's skepticism is: do we need more speed on mobile? In some ways, I share that skepticism. However, I'd also add: do you need gigabit fiber? He notes that he has fiber at home. We often don't know what we'll use speed for until we have it. Mobile data was around for a long time before the iPhone and while some people had found some uses for it, most of us didn't see utility. Then Apple releases the iPhone and all of a sudden the world starts reorienting around data. It can take time for someone to make a killer-app for a capability.

T-Mobile just launched a 5G innovation center to help people figure out what to do with 5G. Sometimes technological capabilities precede someone figuring out what people will want to do with it. I mean, we had broadband connections capable of 1080p video for years before Netflix started streaming. Today, streaming is a way of life.

Bray also says that "I don’t personally know anyone whose life has been changed by 5G. Yes, family members with newer phones occasionally report that the status bar says “5G”, but I don’t hear that they’re having a different experience." A big part of that is that carriers rolled out 5G before most 5G capabilities were really there. If you're on a low-band 5G non-standalone network, it can be hard for it to feel any different. Heck, my device often performs better with 5G off if there's network congestion (since so many people are using 5G in my area and my carrier isn't really good). But that's because every carrier wanted 5G to pop up on your phone even before 5G's capabilities were really in your hands. Carriers pushed things out a bit quickly in part because they wanted phones to start supporting 5G (for when the real capabilities of 5G got deployed) and because it offered useful marketing.

Bray also says that 5G started in Spring 2019. That's not wrong, but it can be misleading. Verizon launched a small amount of millimeter-wave 5G in Spring 2019. Millimeter-wave spectrum typically has a range of 100-200ft and is blocked by almost anything in its path. T-Mobile launched low-band 5G in December 2019, but again low-band 5G in non-standalone deployment isn't really that different. T-Mobile really started launching mid-band 5G in Summer 2020 with Verizon following 1.5 years later in January 2022. However, it takes a while to upgrade a network. If you have 60,000-80,000 towers that you're upgrading and you're upgrading 1,000-1,500 per month, it's going to take a while. For T-Mobile, we know that it's going to be around 3.5-3.7 years to get to 90% of Americans covered. Verizon started deploying mid-band ahead of their launch (since they knew the spectrum they'd have before they were allowed to use it), but they're staying it'll still be 3 years to get to 75% of Americans covered (and given that they started deploying 6-9 months before their launch, we can say 3.5-3.7 years to get to 75% of Americans covered). Basically, we're starting to see some great capabilities some of the time, but we probably have another year or two before it becomes really common and gives people more "wow" moments.

So I think if one is skeptical of 5G, one needs to talk to people that have been using "real" 5G (rather than 5G pushed out for marketing reasons) and one might need to wait a few years as networks get deployed. However, there are already so many people that 5G is making a difference for and that number will grow over time.

That’s a really good point. The first LTE cellphones I had were much faster to connect than 3G when loading webpages, but they weren’t as fast as they got a few years later, as LTE-Advanced started deploying, etc. Although part of the speed difference was probably also due to faster storage and more new LTE bands.

I’d also point out that Canada had one of the fastest LTE networks and still does. Presumably part of the difference is having only a couple dense metro areas to cover. A monopoly on high prices with few independent MVNOs also doesn’t hurt.

T-Mobile’s 5G network has been able to get me off of Comcast and onto their home internet service. Speeds in my neighborhood have been great, typically 300Mbps download / 70Mbps upload.

I’m thankful that this has created competition in the home internet market in my area.

That upload speed is tremendous compared to Comcast's regular offerings. How is your latency compared to Comcast? How is your overall experience compared to Comcast during periods of likely congestion?

Ping times are higher for sure. I’ll try to check what they are when I get home.

Comcast worked quite well in my neighborhood. I always received the speed I was being charged for. The TMobile service isn’t quite as reliable, but it’s still quite good. I spend a lot of time on video calls and never have any issues even during high usage times.

Thank you for the detail. If you don't mind me asking, what pushed you to make the switch?

Excellent detailed write up; thank you very much for this.

>...my device often performs better with 5G off...

Unfortunately, I experience this as well and it's rather annoying. I'm a T-Mobile subscriber and, oftentimes, my experience is better using LTE versus 5G.

The product you are talking about is called “fixed wireless access” and at least in Norway is a huge success. It makes sense because Norway is sparsely populated and this product can serve communities that are hard to reach by fiber at comparable speeds. At the same time it is both better and cheaper than satellite internet, including the spacex offerings. Spacex would make sense for even more isolated communities.

Also, 5G, at least at this point, does not offer consumers anything different compared to a good 4G. But it makes a huge difference for commercial usecases. For example, Norway wants to stop using tetra. Tetra is a network used by police, firefighters etc, which offers robust communication but at a big cost and with very limited services. With 5G they will be able to use the much cheaper public network, with modern services (e.g. video calls), while maintaining robustness and traffic isolation.

One thing I never understood is why it is called "5G" (and not just LTE++)? All the previous "Generations" were pretty big technological leaps in the physical layer, each G was a completely different beast using a completely different physical layer, but from what I gather, "5G" is just LTE with some added frequency bands, bigger QAM, more MIMO, etc (just like LTE to LTE+ added, nothing generational). So, just "before we used to divide this thing into 64 parts and now we divide into 256 so we have 4x the bandwith", nothing that couldn't be done into LTE++ (or LTE+2, or whatever). The only breaking change seem to be the change in the orthogonality of the up-link channel modulation (down-link remains unchanged).

So, can anyone explain to me what is the generational difference between LTE and 5G (not simply more frequencies and tighter parameters)?

5G isn't LTE. It's NR. I'm not sure what the point of questioning the terminology is.

The point of the question is NR is not a generational leap over LTE, it is just a small evolution of LTE (just like LTE to LTE Advance), with the exception of the orthogonality pf the up-link, everything could have been implemented as extensions to LTE instead of a new thing.

I have 5G at my house, I am served by what the provider calls 5G ultra wide band, which is deployed on mm-wave band (28Ghz) and uses phased array technology to implement beam forming, I generally get about 2Gbps on speed tests, but of course through normal activity you can rarely get that to anything on the public net, but to Azure and AWS I can get peaks on certain applications I am working on. The latency variers because of the DRX and cDRX mechanisms but generally I see about 20 - 30ms on start of a flow and 5 - 6 ms during a flow. It will fall back to mid-band (4Ghz) which uses MIMO instead of beam forming due to the longer wavelengths PAA, is not practical, I get around 600Mbps. It will fall back usually on my mobile device even just moving around in the house but the stationary CPE device I have set up is pretty solid. Of course if you connect via Wifi, which I suspect most people would, then you are probably not going to get those speeds but would depend on all the usual wifi caveats. The service is better than the wired service (from a cableco who is run DOCIS 3.1) that I had before

But nothing I have seen yet is in the 10Gig range.

But TBH, I am not sure how most people would use it anyway, at least not today. Maybe there will be some future use cases but today I am un-convinced that some of the hyped use cases will ever pan out. I have read about things like "Remote Surgery" I cannot see any surgeon getting malpractice insurance to cover that. I have the same skepticism about all the various vehicle anti-collision ideas, it seems to be that DRC (Direct Radio Comm) is the only thing that would make sense, the minute you stick a network behind it you are adding a lot of risk of outages and too much latency, I just don't see where you need more computing power in some MEC (Mobile Edge Cloud) than could be put in a vehicle itself. The only that maybe needs that kind of computing power, throughput and low latency is the AR/VR type application which I think is why people seem to be putting 5G, MEC and Metaverse into the same sentence more often lately. My conclusion is that until there is a killer app, 5G will be perceived as a lot of hype. I do not know if Metaverse will take off or not, time will tell.

One of the interesting though prosaic uses of 5G is the ability to replace a bunch of bespoke over their air protocols for "smart" devices. These are boring devices like utility meters, SCADA stuff, and all manner of remote monitors.

Since they can just be "5G" devices they can massively benefit from economies of scale. They can also end up more secure than existing devices because they can more easily be on an IP-based VPN on top of network segmentation/encryption 5G provides.

One of the goals of 5G was allowing low power devices to also have a nice low power radio that can benefit from existing public networks. Not having to maintain custom base stations and relays for some custom radio stack makes a lot more remote monitoring projects economically feasible.

My fundamental issue with any wireless solution for home internet is that ... I have a home. It's not going anywhere. Even with a 9.x earthquake a la Fukushima, the average probability is that it's not moving much in the next 100 years. It's got copper, most of it is for power and pipes. Seems pretty trivial to run a tiny bit more for data. Zero probability that a common atmospheric bus serving the entire area (1) is going to guarantee me as much bandwidth as I can get over that wire. Why engineer freespace optics when a few pounds of circa 1950 oxygen-free copper will do?

(1) area in this case is another definitional headache all it's own, driven by power and wavelength.

5G is weird, since it's being sold as a big deal, when it really isn't.

It has some nice improvements, with the big ones being that it makes more efficient use of the spectrum. I'm not really an expert on this, so if someone is, feel free to correct me, but my understanding is that some of the big changes with 5G NR are making better use of good signals via higher-order modulation, allowing less resources to be allocated to low-bandwidth users, and decreasing latency by allowing communications to start more often.

That doesn't really help any one phone, at least ones that are working well now. What it does do is to make the network better, in the same way going from 3G to LTE did.

Note that their goalpost can always move to "but does it justify the trillion dollar buildout" even in the face of incremental improvements or new use cases

They can calcify their opinions by imagining that they know what a company or ISP needs to do

They've already received many impressions and responses on twitter from their questions about whether 5g has changed anyone's day to day, and did a little survey when working with vendors, and its not likely most people did an empirical analysis on their own day to day to begin with. So it isn't really necessary to add more to that

Agree with author on the last part … I don’t need this. Bought an iPhone 13 mini recently and one of the first things I did was set the cellular from “5G Auto” to “LTE” to save battery life.

My suspicion has been that this tech isn’t for me, but actually more for the companies that are deploying it. For those who know the tech behind this better than me (ie, most of you), does 5G make it possible to turn my phone into a signal booster for my neighbor? Am I now a piece of the network infrastructure with 5G in a way that I’m not with LTE?

I am with Tim Bray in that there is an enormous amount of woo woo spouted about 5G. Perhaps a better question to ask is what is possible now, in our current paradigm, and what might be possible if there was ubiquitous high speed coverage blanketing a complete society. The first is dull, and dubious: a bit more speed. Great. The second is the province of speculation and the source of most of the nonsense.

However, telcos are afraid of missing out on the value chain of 5G like they did with 4G (all that money going to apps, and not a bean to telcos). So they sunk billions into 5g spectrum believing that was the gateway, and they they need to hold on for dear life until those fancy use cases come alive. That could be 1 year from now, ten years or perhaps never. who knows?

In the meantime their investors are watching closely at the only metric they have which is uptake in the consumer domain. However this is weak beer since you don't pay off $70B of spectrum costs $10 at a time (ie consumer behavior along wont bring ROI).

So, pivot to businesses some of which really are groping their way to 5G. Metal buildings, outdoor coverage, military, these are real cases where existing solutions are not great. The gazillion dollar question is whether 5G from a telco is the solution. Maybe its CBRS with 5G, which cuts out the operator. Maybe its wifi 6E, which is fundamentally different from wifi 5 and has learned a lot of lessons from how cellular networks are built. imho there is real risk that 5G vs WiFi is tbe betamax/vhs or bluray/netflix battle of our time.

I have the same sentiment of 5G working better for tethering+zoom calls as https://www.tbray.org/ongoing/When/202x/2022/03/26/-big/cons... - it's noticeable as with my current provider/place I currently have to position my phone in the right area of the house to get 5G.

I'm not on ultrawideband 5G or anything, but it's worth noting that I hardly did video calls before the pandemic and was fairly 5G skeptical before. I later realized video call's value not for dayjob work but for fictive kinship interest groups and study groups going past 'doing conversation' or 'feel-good conversation' but more into support and 'depth conversation'.

I could get by on doing this with a wired internet connection or by doing zoom on the phone (where I think cellular bandwidth QoS prioritization comes into play and Zoom will still work over LTE), but staving off yet another thing to pay for has been nice.

I bought a 5G-capable router because I was moving house during covid and couldn't afford to wait potentially weeks for fibre to be turned on.

It was only supposed to be temporary but it's so good I never bothered getting fibre. While latencies are worse than fibre I can still play Battlefield V on it, and I spend most days ssh'd into GCP machines in various regions.

I can take the router anywhere, and use any network provider I want.

There are features of the 5G network that will enable carriers to provision certain services more efficiently and cost effectively than they currently can on 4G networks, regardless of whether those services (and business models) have any use the for the potential extra bandwidth available. Now whether the carriers will actually launch such services, and whether there will be take-up of them is another question.

This misses the fact that the increased bandwidth allows for improved applications that couldn’t have existed without it.

Without broadband, we’d have fewer applications that could use it. With widespread 5G, we now have the infrastructure (or soon will) for a slew of applications that we haven’t thought of yet but couldn’t have existed without the additional speed.

>although rollback netcode is producing remarkably good results these days.

Tell me you are not an online gamer without telling me you are not an online gamer. Rollback netcode makes about sub 100ms ping playable. Having a low ping is still better than having a high ping. LAN is still the definitive experience since latency is minimized. There are online competitions that are west coast or east coast only because the experience just is not good enough even with rollback netcode to handle latencies that are so high.

When I tried tethering to my phone on 4G I was getting like 150ms ping and it was really spiky. You are not going to have a good time with that much latency.

Also in the gaming context related to low latency is cloud gaming which also includes VR / AR rendering which needs to have an accurate location of where your head is to render the frame correctly.

It also depends on the game. Rollback netcode is for games where your client fully simulates the game. This doesn't happen on a majority of games because of hidden information.

I don't even know if it scales beyond a fighting game, where you'll see rollback most commonly used. Part of the rollback is to go back to a saved gamestate a few seconds ago, and resimulate the entire game with the new (correct) input. Meaning to rollback 100ms, you would need to be able to resimulate that 100ms of gameplay in say 1 frame or 16ms.

>This doesn't happen on a majority of games because of hidden information.

In this case you can just have the server do the rollbacks as it can view all of the information.

>where you'll see rollback most commonly used

Personally I have run into rollback in FPS games. Not many fighting games use rollback.

I think this post is sort of looking for a "special reason" for 5G to exist that actually doesn't need to be there for 5G to completely alter the telecom landscape.

Bray touches on this a bit, but at a personal level the biggest difference for me is that, at home, my T-Mobile 5G connection is consistently faster and more stable than my AT&T UVerse connection. It's close to the tipping point where I'm not quite sure I need cable at all. That wasn't the case when I was on LTE.

All the other marketing BS that carriers and equipment providers can dream up may just be bullshit, but switching to a world where many consumers could get all of their telecom and TV desires met from their cell company instead of their cable company would be a massive change.

The author lives on a boat.

Imagine going back thousands of years and hearing people that lived on land saying "there's no need to live on the ocean, the land is perfectly safe and there's just no good reason for it".

My stomach turns when people are against technology for reasons other than ethical ones.

Um I use our boat as an office, bicycle down to it most days. Ain't spacious or luxurious, but it's waterfront. I live in an ordinary house with decent (~300M) wired connection.

There's some classic Tim curmudgeonry right there! Is he complaining about the hype or the tech?

There's a continuum between "5G is an unneeded upgrade pushed by [insert conspiracy theory here]" and "5G is an essential new invention that will significantly improve our lives." The truth is somewhere in the middle, and in my opinion, leans much more towards the latter.

I was pitching Vodafone business customers in Spain on the power of the coming 3G revolution back in 2001. The stuff we were selling as "coming next year" wouldn't be commercial products until over a decade later, and required the invention of the iPhone in between.

That's just how it has worked for decades now. This isn't a surprise.

> ”Question: How often do you need more than the 50M or so LTE offers in a situation where it’s cheaper to provide it with 5G than with a wired connection?”

Here in the UK, you can get very fast 5G for significantly cheaper than a wired connection. I’m on the Vodafone network and get 300+ Mbps for £30/month. My mate just got a SIM on the Three network which gets up to 800(!) Mbps for £20(!)/month - although his has a 12 month contact.

Plus, many 5G routers are tiny and have batteries in them. I can pop mine in my backpack and have super-fast internet wherever I need it.

Fixed wire connections typically come with long contacts, they don’t move with you, they cost more and they’re less reliable. They do have lower latency however!

I've never noticed an improvement from 5G over LTE.

What I care about most is reliability of some low but usable amount of bandwidth, as opposed to maximum bandwidth. Both 5G and LTE are annoyingly unreliable for me (wandering around in Silicon Valley).

Yeah, going for a walk round Seattle on a VoIP conference call on either 4g or 5g, and both drop out for half the trip

5G on higher frequency could be less stable while moving.

For me the problem is that you often can't even tell if you are really on 5G. Some carriers label certain types of 4G as 5G, and my understanding is that even if you have "real" 5G, it probably isn't mmWave 5G in most areas. So even if you are sitting there with a good connection on a mobile device that says "5G" in the corner, it's hard to say that 5G is really making a difference.

For example, I'm sitting here in my car far away from Wi-Fi. I just did a speed test with 5G on and off. On 5G, I got 63mbps up/16 down. With 5G off, I got 100mbps/25mbps (much faster).

In terms of increased thoroughput, hz for hz, compared to 4G LTE, 5G NR can send about 15-20% more data. That's not a huge improvement compared to past increases of multiple hundreds of percent (3G-4G). Because of this most increases in thoroughput have to come from using new frequency ranges. So they telcos pretty much stole half of the 3 GHz C-band from incumbent satellite operators (who were told to "just use h265"). But beyond this there's not much available spectrum that's even just decent like C-band. Most of it is crappy mm-wave.

One one hand it is annoying to lose that type of spectrum. On other hand terrestrial can mean smaller cell sizes thus much better utilization rates. Which really is needed if demand continues to increase.

Yup, there's only 1Mbit/sec for every 1Mhz of radio bandwidth, and this is shared between all users in a given cell. This is also why I don't understand the fascination with Starlink, which makes cells obnoxiously large (the size of a satellite coverage), even Musk himself says it's only for remote areas.

And why I have wondered how does the math really work out. Fiber and cell tower cost only so much. And outside some mobile cases does areas where this isn't viable have enough customers?

Then again, USA is always an argument where telco seems to have really failed for various reasons. Still, I see no reason why same would apply to Africa or other markets they talk about...

> there's only 1Mbit/sec for every 1Mhz of radio bandwidth

That can't be right, cellular networks have had > 1 bps/hz efficiency for decades.

https://en.wikipedia.org/wiki/Spectral_efficiency

> there's only 1Mbit/sec for every 1Mhz of radio bandwidth

> this is shared between all users in a given cell

Both statements are untrue.

For example, "Wifi 7" 802.11be reaches up to 125 Mbit/s for each 1 MHz of radio bandwidth in perfect conditions, to a single user.

Inside a cell area, multiple users can share the same radio frequencies at the same time using various spatial modulation techniques (typically called beamforming but it doesn't have to be a beam, and might involve 3 or more stations working together).

The theoretical limits scale very differently than common understanding of bandwidth.

> reaches up to 125 Mbit/s for each 1 MHz of radio bandwidth

I see. What is the theoretical limit of bits/second per hertz of radio bandwidth, if any?

Ignoring quantum effects, there's no upper limit in theoretically perfect radio conditions, even for a single radio channel, where there is absolutely no background noise or interference, and the antennae and electronics are literally perfect. However, nothing can be built like that, and there is always background noise.

The real life limit is complicated. For each single radio channel the limit depends on other characteristics of the channel than just the (temporal) radio bandwidth, for example the antenna, background radio noise, interference, and power level.

The Shannon-Hartley theorom covers capacity of a single noisy channel with fixed characteristics; it depends on the signal to noise ratio as well as the bandwidth.

Radio channels used for mobile communications are typically "fast-fading", which means the conditions keep changing, and both ends have to continuously measure those conditions and adapt dynamically how they encode and decode the data. This complicates a measure of the theoretical limit, and also means the limit is changing with time, for example as the user moves the handset around.

https://en.wikipedia.org/wiki/Channel_capacity#Channel_capac...

The highest rate encoding used in 802.11be is "4096-QAM" which carries 24 bits/sec per Hz when noise is sufficiently low, and decreases the more error correction is required. I'm not sure if 4096-QAM is chosen because the signal-to-noise level is such that a larger number wouldn't improve thoughput enough to matter, or if it's limited by the electronics, non-linearity and calculations at those high speeds.

That bit rate is for a single radio channel. Multiple channels overlapping the same frequencies are also possible, as long as the antennae are sufficiently spread out in space and/or polarisation. When something large enough like a laptop (which has multiple antenna in the display) talks to a base station, it is combined with beamforming and diversity, or in a general mathematical description, polarisation modes and spatial bandwidth. That is equivalent to having multiple radio channels using the same frequencies to the same device, but the channels are not physically distinct, they are combined and separated mathematically with linear algebra.

As with MHz bandwidth, spatial and polarisation are also limited. You can't add more channels with no limit to a single device of fixed size. But if you have no limit on device size so can add more antenna spread out, and/or many base stations cooperating, you can increase the total bits/second by using those. 4G LTE does the multiple base stations thing to some extent already, to maintain channels to multiple user handsets at the same time inside a cell. It's not that each handset talks to a separate base station. It's that the signals are combined with linear algebra in such a way that each handset can decode different information from the multi-spatial-channel signal that they all receive slightly differently.

Thank you for the detailed response! I just recently picked up an RTL-SDR, and for a moment though that I have this bandwidth thing figured. Nope, more learning ahead. I'll probably write some simple demodulator to better understand this stuff.

PS. I think I can see how to send unbounded data rates through a narrow bandwidth, assuming perfect conditions with no noise. If you want to transfer N bits per second, for half a second transmit plain carrier, then for half a second transmit carrier multiplied by x/2^N DC, where x is the sequence of bits to be transferred in that second. Then repeat the process. The receiver can establish the max reading when pure carrier is being transmitted, and do a readout when the multiplied carrier is transmitted.

That's exactly right, although if you take the Fourier transform of that modulated signal you'll see that it uses infinite bandwidth due to the square-wave modulation steps at each half second.

That can be sent through a narrow bandwidth channel, which will filter out the higher harmonics. But then, as the Fourier transform suggests, you'll see at the receiver a different signal, with rounded edges instead of steps in the modulated amplitude.

Those half second intervals set what's called the symbol rate to 2sym/s, and the Nyquist frequency for them is 1Hz.

If your channel bandwidth (the filter) is > 1Hz, the information will still be there in the modulated, filtered signal to recover, but you'll have an interesting time doing it if the bandwidth is close to 1Hz.

If your channel bandwidth is < 1Hz (the filter strictly removes all frequencies above 1Hz), you won't be able to recover all the transmitted bits.

I recommend building a demodulator and varying the channel bandwidth to get an intuition for how it gets trickier to demodulate close to 1Hz. Note that making the strict bandwidth filter is also quite an interesting challenge, especially close to 1Hz; this is called a brick wall filter. Doing the Fourier transform on the entire signal before and after filtering is a good way to visualise the bandwidth taken by a modulation scheme and the effects of a channel, and to see if it's correctly removed everything above the threshold frequency. Modifying those Fourier values in the frequency domain then using an inverse Fourier transform to construct a filtered version of the entire signal is a good way to construct an ideal filter for testing the demodulator.

I phoned 3 and complained that my 4G internet in central London was so terrible that I might switch - it was going down to 0 bits per second so you'd fail to call an Uber. They said, sorry they were installing more gear for 5G and things would improve and they have, although I'm still on 4G. I think the networks constantly upgrade with faster and higher bandwidth stuff and the terms like 3G, 4G and 5G are mostly for marketing. So I'm all in favour of 5G upgrades but feel no desire to get a handset that has '5G' printed on it.

My takeaway from this is pretty simple and perhaps the most boring answer but: it depends.

The biggest thing the author seems to say is that what we already have is “good enough” which might be true for them but isn’t applicable to many other people or applications.

The technology works better or worse depending on how it’s deployed. Not a huge shocker. Deriding the marinas Wifi doesn’t inherently mean it’s bad or good as a technology.

5G certainly oversold as revolutionary but it’s fine And has a higher upper bound for obvious reasons as a technology.

5G might be great for some people. My experience with it is that in the places where I used to have poor or no connection I still have no connection. The places where I used to only have a 3G connection I now have no connection at all. The places where I get a 5G connection my phones swaps back and forth between a fast 4G connection and a pathetically slow 5G connection which repeatedly interrupts my data. I modified my settings and turned 5G off entirely. Good riddance.

Okay, 5G is obviously better than LTE for people who needs a better wireless experience, or for people who does not have wired connections.

But my question is, is it significant enough to change industries? In case of LTE, we finally have reliable bandwidth to watch Youtube and Twitch. Even those social media have existed before LTE, but I believe they exploded since LTE.

Then what about 5G? A lot of marketing said 5G is needed for VR, MR, AI, Autonomous driving, and etc. But did they really need 5G?

I am using a 2015 Nokia Lumia 830. I have been using it over the pandemic as my work modem taking video calls. LTE is great with multiple computers paired. 5G is overrated marketing garbage. Especially on the Galaxy Fold 3 on t-mobile. Where it keeps going in and out.

All this obsoletion of old devices just so the networks can keep pushing more marketing data and sell more shiny. At the cost of environment.

Interestingly (to me), I just turned 5G off on my phone yesterday. I’ve had weird connectivity issues ever since I got this phone, even when I see “5GUC” (“ultra capacity”). Out of idle curiosity I wondered if LTE might be less spotty. So far it mostly has been! But ironically it cut out in the course of writing this comment, while looking up “5GUC” to confirm I remembered what it meant.

I tested LTE in my city and the places I go I range from 20 megabits to a smidge over 100 megabits. Honestly, for what I do on my phone while I'm out and about 20 is perfectly adequate. I simply do not need faster. The only benefit 5G could bring me that I see is maybe less congestion. But ... I don't really see and issue with that either.

This article, like most human opinions, is selfish. It disregards that other people want 5G. It concerns itself with only a personal preference of one human being who lives in an area where population density is probably 5 per 100 square miles........ You can't even send a photo when connected to LTE at a crowded NY International Airport,even though the signal is strong,five bars, and you're connected but there are also other people connected. It took me 20-30 minutes to send a photo via Viber last time. Yes sure they could build a second tower and use 4G but 5G might solve that with one tower and less money, it's also better........ Yes 5G will probably make more money and that's great, if it didn't make money it wouldn't make sense, and in the process maybe next time I'll be able to send a message and maybe people who want more than 20 Mbps can have more.......... It's very selfish to assume the purpose of building wireless communications is to watch football games and Netflix, while that is one purpose for some people, other people use it for something else. There are those who hoard data, I'm not one of them but I do recognize, people who work from home and need lots of speed and reliability etc.......... I would understand if this article had an argument such as "5G is unhealthy" but it doesn't. This article's argument is "I don't know why they would need 5G as 4G works fine in Canada" and then it implies people shouldn't progress on the mere argument that the need for progress isn't understood by the writer. I don't claim to understand the need for 5G but I acknowledge that I don't possess that knowledge, the writer doesn't acknowledge that he does not know why those people are building 5G. He basically argues 4G works fine for me therefore all the engineers should stop working. ....... One shouldn't make an argument on what other people should do based on personal situation and own opinion disregarding what they want and think, unless 5G causes pollution or health hazard. Let it be. ....... Yes the writer might think the word sceptic might have masked the implications of the article, it didn't. Also word sceptic doesn't actually mean what it's euphemism means today. A sceptic is someone who thinks, and here the writer only thought his own thoughts, didn't think about what others need and might think, so not a sceptic at all, in this article at least. ....... Opinions are great when they include other people, when they include only the self they're just selfish.

It's strange how everything related to to mobile networks is meh or bad in the US.

tbray and others think it's the technology they are analyzing when it's actually how operators adopt the technology that makes the difference. It must be market power issue. Operators just do the minimum possible coverage they can get away with.

> It's strange ...

Not strange at all. They're "job" is to funnel as much money to investors as possible. They can all get away with being awful because there's only 3 of them left.

Yay capitalism?

> It's strange how everything related to to mobile networks is meh or bad in the US.

Hasn't been my experience. Over the past decade or two, coverage and speeds have increased a ton and prices have stayed low. Carriers are spending billions on new spectrum and an ever increasing build out of cell sites.

What are you comparing against?

Compared to northern Europe and rich East Asian countries the US mobile infrastructure and prices are bad in my experience.

Most of the coverage is going to be 800/700/600 Mhz stuff that will have the same problem of peak hours congestion as LTE. The 5G pipedream is based on an assumption that telcos could possibly afford to litter every inhabited place with shitload of HF base stations

By the way, HF does stand for high frequency, but it's actually the 3 to 30Mhz band.

5G mmWave would be the SHF and low EHF bands. (Super high/extremely high)

5G also provides better spacial efficiency over LTE besides the spectral efficiency. So even without a forest of towers it'll provide better peak congestion performance than LTE. At the same time microcells/nanocells are more practical with 5G allowing companies to increase capacity in especially high traffic areas. For instance a couple high-band cells in popular public spaces. With LTE the only equivalent option was a lower power microcell that affected and was affected by nearby full sized cells.

There's plenty of marketing dreck and hype around 5G but there's also a lot of good engineering and real capability improvements over previous cellular systems.

Anecdote that Tim is asking for: I upgraded from an iPhone 2020 SE to an iPhone 12 and suddenly it was easier to use the BlueBike app in New Orleans. With just LTE or whatever it’s called, sometimes the app wouldn’t connect at all and so i couldn’t park my bike.

I used my iPhone 12 as a 5G modem while working from home for a couple weeks after I moved last year before I had my gigabit RCN service setup. It was remarkably performant and stable. I could probably use it full time if it weren't for the higher cost.

I found this article to be convincing in-favor of 5G. I think the real concerns about 5G are how it can be abused by providers and governments to spy on people, none of which was mentioned in this article.

I would find this compelling if I hadn't heard this argument many times before.

"What could we ever do with 1GB of storage?"

"What could we ever do with 1GB of memory?"

"What could we ever do with a much faster internet connection?"

etc

One of the big arguments I've heard from an AT&T PM who is supposedly In The Know is that 5G lets First Responders have privileged bandwidth in a saturated area.

First responders already have their own band with 20MHz of spectrum: FirstNet

In fact, its run by AT&T. Its not a 5G specific thing.

Sorry, does the implementation of 5G take away frequency bands from 4G and 3G?

That would mean a form of planned obsolescence, making 5G faster, by making the older tech slower.

I found this hard to follow so stopped reading.

> [Note: When I say “G” or “M” I’m talking about Gbits or Mbits/second.]

> Why 5G? Faster connections — there is talk of 10G!

Wait, does he think 4G = 4 Gbit/s, and 5G = 5 Gbit/s? If there's talk of 10G well, what about 6/7/8/9G? G is generation. Additionally:

> The phone calls what it sees “LTE+” (I don’t claim to understand what that means)

It isn't hard to find out information. A quick search leads to https://en.wikipedia.org/wiki/LTE_Advanced.

Agreed.

This was a hard to understand article because of that note. The author appeared to switch between G as in Gbps and G as in generation multiple times without clarification.

Would have been way more clear if they just used Gbits instead of G when they meant Gbits. (or Gen instead of G).

author invented xml. while i haven't any love for xml i assume he is on solid technical footing and just worded poorly ?

Oops haha you're right, "G" has two different meanings in the article. Couple proofreaders missed that too. I think most readers are smart enough to figure out which is meant most times?

Whoever did the dsp behind 4g is a genius

Reply on news.ycombinator.com