Ask HN: Is it just me or is 5G strictly worse than LTE? | Hacker News Reader

Ask HN: Is it just me or is 5G strictly worse than LTE?

295 pointsapitman328 comments

I feel like signal is lower in general, and I run into a lot more cases where connections seem to hang entirely. In theory 5G should offer a lot more bandwidth, but I don't remember ever being bandwidth-limited on LTE. 20Mbps is plenty on my phone. Same for latency. I'm not playing FPS games on my phone.

5G seems worse, but I used to work on a physical layer wireless technology and it might be placebo knowing that the higher frequencies used for 5G don't go as far or penetrate as effectively.

View on HN

Having the 5G logo on your phone doesnt necessarily mean your phone is speaking 5G. It's...complicated.

- Most US operators do not have ubiquitous 5G radio coverage, so, they do what is called 5G NSA where your traffic ultimately ends up veing processed by the 4G core infrastructure that they have. Which is already overloaded and a bit creaky.

- Some operators can do real 5G, "5G SA", where the whole flow end to end runs on 5G infrastructure, but whether this is faster than LTE or not depends on the spectrum band in use. Verizon have some high bandwidth spectrum, but it doesnt propagate as well as lower (and slower) frequencies like what Tmo has. But if you expect to get those blazing multi gig speeds, you really have to be on that high bandwidth stuff. For most people, most of the time, they wont, so they wont see much improvement over LTE.

- I assert that it is dawning on operators that consumers are not interested in paying $10 extra per month for 5G. This is a bit of a problem when those same operators spent or borrowed billions to obtain the spectrum in the first place. Did I mention that the era of cheap money is now over and those debt payments are due?

- Telcos desperately need a killer app or use case that drives 5G adoption. They havent got one. And remember, the app must be one that telcos can monetize. They still have scars from their failure to capture the value of smartphone applications in the LTE era.

> They still have scars from their failure to capture the value of smartphone applications in the LTE era.

That's makes a lot of sense and truly, fuck telcos.

Having worked with telcos in multiple countries, I can assure you, we cannot have a lot of nice things because of them.

While I am not big on defending telcos, let's not forget that they are under enormous price pressure (for the fibre backbone essentially they need to double available throughput every 18 month or so without increasing cost) at the same time it's the software/service companies who really made enormous profits based on these developments, while data capacity is becoming more and more a commodity.

man, can you imagine if we had payed some telcos money to drop fiber that they never did?

Still waiting on Centurylink (or whatever they rebranded to now) to start on their piece of that "bridge to the 21st century". Not holding breath. 3Mb/.5Mb, at a cool $50/month. :-O

When they figure a way to bill users differently between "long-distance" and "local" packet transmissions, they can start to see applications/services put in effort to adapt.

Also, mechanisms like L4S for high-fidelity ECN feedback allow for higher link utilizations without end-to-end QoS impact.

IPv6 reducing routing table fragmentation should enable reasonable prefix based peering, along with billing senders according to the used peering. We don't need fast flat rates, we need cheap bulk traffic and flat rates for effectively-not-oversubscribed bandwidth. Maybe with a way to roll over unused priority bandwidth at some multiple into the next billing period to count against that period's bulk traffic usage.

That is because data capacity is a commodity, and should earn profit margins of near zero, like any utility does.

Also, no one asked ATT to lose billions of dollars by buying Time Warner HBO and DirecTV. Or Verizon with Yahoo.

But they wanted to chase that Comcast vertical monopoly dream. And it will never yield any utility for customers.

It should all be a utility, like water, gas, and electric. And do the one job, installing pipes (and upgrading them) to people’s houses and delivering the goods.

imo Telcos have delivered jack shit in terms of "developments" except extract rent from their monopoly positions.

> Telcos desperately need a killer app or use case that drives 5G adoption. They haven't got one.

* Self driving cars? No.

* Internet of Things? No.

* Metaverse? No.

* Remote surgical operations? You gotta be kidding.

Everyone in a big stadium can watch the show on their phones, though. If someone was willing to pay for the hundreds of tiny cell sites required.

> If someone was willing to pay for the hundreds of tiny cell sites required.

We are already doing that, it's called WiFi. 5G SA is nothing but a souped up WiFi signal, and it should be treated as such and not how we have been treating 4G up until this point. Telcos need to change their business model otherwise the tech giants might end up eating their lunch

The tech giants spy on us to sell advertising inside of their products. If Verizon starts listening into my calls to sell me products I would switch to any carrier who wasn’t doing that.

This sounds very familiar. I hope they get a hundred billion dollar fine and life in prison for the CEO and the board if they do this because they have already been fined before for this.

> Verizon is settling with the FCC over its use of an ad targeting technology known as a "supercookie," which tracks the websites visited by phones on its network. Supercookies allow websites to better target ads to visitors with Verizon cellphone service; but those visitors — for a period of time — weren't informed of the tracking or given the option to opt out. Because of that, Verizon will pay a fine of $1.35 million and will now have to receive customer permission before sharing tracking data with other companies or even within its organization, including with sites owned by AOL.

- The Verge

Underlying FTC press release

Full Title: FCC Settles Verizon "Supercookie" Probe, Requires Consumer Opt-In for Third Parties

https://www.fcc.gov/document/fcc-settles-verizon-supercookie...

> WASHINGTON, March 7, 2016 – The Federal Communications Commission today announced a settlement resolving an investigation into Verizon Wireless’s practice of inserting unique identifier headers or so-called “supercookies” into its customers’ mobile Internet traffic without their knowledge or consent. These unique, undeletable identifiers – referred to as UIDH – are inserted into web traffic and used to identify customers in order to deliver targeted ads from Verizon and other third parties. As a result of the investigation and settlement, Verizon Wireless is notifying consumers about its targeted advertising programs, will obtain customers’ opt-in consent before sharing UIDH with third parties, and will obtain customers’ opt-in or opt-out consent before sharing UIDH internally within the Verizon corporate family.

Verizon may not listen to your calls, but they do track the IP addresses your phone connects to and what you do on the internet to profile you.

Source: PII data request of my own Verizon account, containing detailed demographic data, intent/interest data, and IP addresses of websites I had visited.

Edit: Anyone interested in doing the same, here's the link: https://www.verizon.com/support/download-and-view-vpd-file/

WiFi is a 'simple' system designed for a few, static devices.

Cellular systems are designed for many more devices and, crucially, full mobility. These are orders of magnitude more complex than WiFi.

Then, specifically 5G is about increasing capacity (number of devices), reducing delays, and stringent reliability.

There's nothing within the 5G spec itself that commands that every single cell site needs to be a massive MIMO supporting 1000s of users at once. Making a micro cell supporting just a family home is enough, then something a bit higher for small businesses supporting 10-100 users, then the likes of Starbucks can deploy bigger ones and finally the Sports stadiums and Corporate offices gets the really high end stuff.

Companies like Rakuten have already shown it's cheap enough to distribute 4G femtocells for free [1] to your users while increasing your coverage

[1] - https://network.mobile.rakuten.co.jp/guide/rakuten-casa/conn...

Again, 4G/5G are not WiFi, they are much more complex and reliable systems and support mobility, even if of course they can now be used in the same way as WiFi ("he who can do more can do less").

T-Mobile continues to quietly make available their "CellSpot" femtocell for a US$25 deposit, which is good because it's the only way we would have actual cell service here in a rural part of New Mexico. It is doubly good because whatever server T-Mobile uses for their wifi calling service(s) is the biggest dumpster fire in their entire business.

I get fantastic wifi calling service with TMO. Maybe your home internet has latency issues?

Non-TMO VOIP and video calling works like a charm. By contrast, TMO Wifi calling can't deal with MMS SMS, drops calls and generally is flaky as shit. The cellspot makes all of that work just fine, and is better than driving 2 miles to get to a reception point.

> 5G SA is nothing but a souped up WiFi signal,

Let's not conflate too many things. Maybe current deployments are in that frequency band, but "5G SA" just means the connection is legitimately 5G.

In theory all 4G frequencies will get converted to 5G, and it will be a moderate improvement. Separately 5G is expanding into the wifi arena and the ultra fast near-line-of-sight arena. But "SA" is not how you tell the difference. It will all be SA eventually.

> Separately 5G is expanding into the wifi arena

On LTE this is called LAA. As far as I can tell, the term is being kept for 5G as well.

I worked on a team that prototyped the “watch the game on your phone at the game” using WiFi. Nobody used it. That isn’t going to be the killer app either.

I would expect that to work great for a F1 race.

It working great and people actually wanting to use it are two different things. In my experience people go to sports events to experience the atmosphere and enjoy themselves while having a beer, not to look into every minute detail of what's happening.

Exactly... the most fanatical football nuts I know are the ones watching the game on TV with the different angles and the slow-motions and the replays... the people going to the game are very tribal, they love painting their faces and shouting [CITY NAME] and they do genuinely get very happy when [TEAM] wins but they aren't the ones talking about the 0-technique of the defensive lineman and so on.

Of course, but having been to an F1 race recently, the tracks are so irregularly shaped and long that it's usually impossible to see the whole track (or even more than some small section of track) from the stands. And even the big screens can be sparse.

It's been tried, and the results depend on your definition of "works great".

F1 used to have a deal with "Kangaroo TV" [1], where people at the race could rent a small portable TV. When I went to the Shanghai GP years back (it was the one where Hamilton beached his car coming into the pits), I rented one. With racetracks that are miles long so nobody can see all the action - and certainly not when sitting at Turn 1 in Shanghai - it's definitely a boon. I was able to see what was happening, even listen to the commentators. I was telling everybody around me what was really going on.

The thing is, it was fairly expensive. I don't remember the actual number, but something that you probably wouldn't just do on a whim.

And history tells us that it didn't work out. The "kangaroo.tv" domain name is apparently available for purchase now.

Too bad, because my experience is that, without this, going to an F1 race is inferior to watching on TV in most respects.

[1] https://www.autosport.com/f1/news/fom-joins-forces-with-kang...

Livestream yourself watching the game / gig might be a thing though. I went to a Phoebe Bridgers gig in the summer and half the crowd seemed to be at it. Not very good for the atmosphere though.

What if players had first-person camera streams available, and you could choose which stream to watch on your phone?

Works great in F1, would be fun in soccer, football, or even basketball.

IoT is extra funny because the devices are so low-bandwidth.

The point of 5g for iot is a special mode with very low battery usage, and a way to connect without needing a SIM card or something similar.

There's value there, but not for consumers. It's large land monitoring where this gets value. Essentially competing with LoRaWan by building out a ubiquitous radio network any sensor can easily connect to.

I think the average person just doesn't realize how much remote sensor equipment infrastructure is out there. Some number of years ago I went out with a friend that worked at natural gas well facilities. I'm pretty sure it was all over lower bandwidth radio networks at that time, but it still transferred all the important data they needed. Such as waste water tank levels, average gas flow, and if any emergency or alarm conditions were occurring. More monitoring and data was always wanted but you had to keep it within particular power and bandwidth limits at the time.

5G highspeed signals are so high frequency that they penetrate no kind of matter anymore. Good luck. Only usecase maybe the stadium with a central antenna.

I live in a poured concrete house built in 1950s Ireland. My 2.4Ghz doorbell struggles to speak to an AP about 1 meter away from it through the external wall. I just did a 5G speedtest from my bed and got 250Mbps.

And which 5G it was? Because mmWave one won't really penetrate anything on 10GHz+

They are certainly not using mmWave 5G. Numbers show (1st Google result) that less than 1% of time connected is to a mmWave network. I don't personally get why mmWave part of 5G is talked so much about.

5G over the same cmWave frequencies 4G uses still has a benefit - but mainly for the operator! It allows them more users and throughput for the same amount of spectrum. So in a rational market, the operators should be giving you discounts for using 5G instead of charging more for it!

I just benchmarked from my suburban second floor bedroom in the US, and got mmWave (it says 5G UC which is apparently mmWave) and got 500mbit/sec

So I guess some people have it, and use it? It’s basically always on around my neighborhood, but I guess I live in a nice city and Sprint’s old HQ is down the street so that probably has a lot to do with it.

That is is T-Mobile “Ultra Capacity”, which is more like.y the midband spectrum they got from sprint.

You basically need line of sight for mmWave. Think of it like light. If you can’t see an access point, then you probably aren’t using mmWave.

Half a bit per second?

Only the US really uses mmWave.

~3.4GHz-4GHz (which is the 'primary' 5G band) in Europe can provide excellent speeds (2gbit/sec real world no problem to a phone) but without the terrible signal degrigation you get on mmWave.

mmWave will be used more in the future once the other bands get saturated, but it's a bit ahead of its time now.

It's probably coming through your window, not the walls. Try it from a bathroom.

Thankfully, I spend virtually all of my time indoors in rooms with windows.

I know people who are implementing server-based "responses" in cars (e.g. tyre slip control - similar contexts) and need the lowest possible latencies for that.

Of course, some of are of the position of "You try to place remote access, control and delegation to critical pieces of my vital infrastructure, expect the most violent level of personal reactions".

Source?

I mean. We’ve all heard the marketing bs from the telcos that 5g opens up these possibilities. No engineer, worth the title, would be naive enough to actually believe that and put tire slip control into an rpc call...going over wireless. For real, does anyone actually believe this is a good idea?

There are things worth putting on remote systems, but none of them need real time response or matter for safety.

I work for John Deere (of course do not speak for them), we want to analyze things 'on the cloud', but we are thinking about how to apply the results of data we gather latter. Already it is so the coop sprayer that comes in 2 months knows where the rows really are so it can figure out the most efficient path to drive their field and drive over the least crop. There are also what do I do in this field next year decisions that can be driven by data, and it is nice to not have to take some form of memory device to the tractor.

Companies with trucks are interested in things they can get from connected vehicles. Insurance also is interested to see if they can find the good drivers and give them better rates (governments probably want to force training on bad drivers or something). None of the above is real time critical.

> Source?

Impossible (unfortunately) for reasons of privacy.

> No engineer ... naive

Your mistake is to imply that the "drive" lies with engineers - instead of a perverse economy in which some try to promote unneeded novelties to "have something new to sell - to people who already have everything". This was directly admitted by the involved.

> a good idea

Given the above, some believe it a potentially good idea "for the stockholders" (not for "function", as you seem to assume as a perspective).

I thought the killer app would be to replace at home wired broadband with 5G wireless. I still think it looks compelling.

Maybe in the developing world but I'd imagine most of the developed world got rid of wired internet and replaced it with fibre years ago.

Where I live we have decent fiber penetration. In my appartment I can have a 1Gb/1Gb internet connection no problem.

However, I pay about 25 eur for my 5G telco (unlimited data), and 35 eur for my fiber connection (100mbit/100mbit). My telco allows dual SIM usage, so I am thinking about just cancelling my fiber connection and installing a 5G modem. The thing that is holding me back is that a good 5G modem is somewhere between 500 and 4000 eur, so it is quite an expensive experiment to do.

The irony in that is that it's probably the other way around but I'm too lazy to dig up data on that.

Outside cities, the network situation is pretty bad. I am in an affluent suburb and my only option is 4mb/200k. No that is not a typo. Broadband coverage is still pretty spotty in the usa.

Wow, we have fibre to about 90% of the population and ripped out copper a few years ago. In rural areas there is government subsidised 4g internet and in the most rural it's government subsidised satellite.

I was in Korea for work and it seemed the same there, I was even in eastern Europe visiting family and it was a pretty similar story. So I figured it was pretty universal. Although I know that Australia isn't in a good space with broadband simply because of politics and companies not wanting to work together, I'd assumed they were an outlier

You wrote:

    we have fibre to about 90% of the population

Australia? NBN (National Broadband Network) is the only one I can think of...

Real Question: On paper, NBN is an insane human development achievement. Why do so many non-urban Ozzies on HN complain about lousy Internet access? Honestly, I don't believe it. Send them to Germany for a year, and they will see how bad it can be. (HN is full of funny stories from Germans about how bad is their broadband.)

No, New Zealand with UFB which is fibre to the door for about 90% of the population with rural broadband initiatives for the remaining 10%.

Trouble is, as far as I can tell, at 5G frequencies you need micro stations all over the place. If your suburban street has 10 houses and needs 5 stations (with fiber runs to each) to offer mmWave to everyone, it’s basically just the same thing as home broadband yet more expensive.

Nope. There was a company about 20 years ago called WinStar that tried this. The reason it’s not “basically the same” is called “the last mile” problem. It’s vastly cheaper to install things on every block even than to every house especially when the utility poles and right of ways are already there.

At mmWave 5G frequencies. 5G also runs on 4G frequencies with the same propagation as 4G.

In something of a paradoxical curse, operating on 4G frequencies mean you share spectrum with a much larger land area, limiting possible bandwidth. You aren’t going to serve 1gbps of home internet to every house in reach.

I thought this, too, but it seems carriers in the USA are being /very/ selective about where they offer the service.

For example, my inlaws live in semi-rural South Carolina and a new cell tower was built about a mile away from their home ~2 years ago. T-Mobile's Home Internet product indicates there's no service at their address for even though I have what looks to be full signal strength indoors on their "5G UC" network (indicates mid-band & mmWave from what I can tell, but in this location it's likely just mid-band). Every speed test I've executed from my T-Mobile-connected phone (Galaxy S22) is in the 600-700Mbps range and it doesn't vary based on the time of day.

Verizon Wireless indicates that they're not in the service area either for their 5G home Internet product, although I don't have data points from any UE on their network in that area.

Verizon even markets "5G Nationwide" as "performance comparable to our award-winning 4G LTE". So what's the point?

You have to spend more $$ to get "5G Ultra Wideband" on your plan, which is the real deal.

Wow, they actually said the "quiet part" out loud.

> - Most US operators do not have ubiquitous 5G radio coverage, so, they do what is called 5G NSA where your traffic ultimately ends up veing processed by the 4G core infrastructure that they have. Which is already overloaded and a bit creaky

Interesting! The first time I ever remember my phone claiming it had a 5G connection was late 2020/early 2021 (I forget the exact month, but it was within a few months of me moving to a new city). I had taken an Uber downtown, and when trying to do the same to return, I was surprised to see that I couldn't get anything to load, and my phone claimed it had a strong 5G signal. I couldn't find any way to tell it to try to use LTE instead, so I ended up having to walk maybe 10 minutes back in the direction of my apartment before my phone finally could load anything, which coincided with it showing LTE and not 5G.

never knew exactly what had caused the issue, but I suspected it might have something to do with a cell tower reporting 5G to my phone but there not actually being infrastructure to support that in the area. I wonder if something similar to what you describe was happening, and due to an OS bug or something my phone had different expectations of what communication would look like when the other end reported that 5G was supported. Does "5G SA" require any different behavior on the phone's side of things, or would any communications it receives be indistinguishable from those of "proper" 5G?

The “real” 5g isn’t for your phone, it is to replace your home internet.

For that, you want fast and low latency and that means a dense amount of cell towers. The fact that “real 5g” has a shorter range is actually a positive. It means you can make your shit super dense.

It's almost like we could just go back to cables and have a plenty good experience again...

But why? If you had the choice of EQUIVALENT (bandwidth, latency, jitter, reliability, etc…) network connectivity, why pick a wire over wireless?

Wireless consumes more energy. So it is more enironmentally friendly to go with wires, especially fiber. Personally I like to have both, wired main connection and wireless as backup.

Is it though? What’s the environmental footprint of manufacturing, installing, and maintaining all of those wires?

While not as much, installing and maintaining the 5G endpoints also requires all that, and they generally have to be wired to achieve low latency, meshing only does so much and the speed of light is going to win out for backhauls.

A lot lower than manufacturing chips to do the complex signal processing needed for fast and reliable wireless.

If it costs more, I wouldn't pick wireless. Losing the wire for my home internet gains me nothing.

I don't trust that the wireless solution would be equivalent, though this may be because I've been burned too many times living in concrete houses that block cellphone signal rather effectively.

It isn’t equivalent. Optical fiber beats 5G on those criteria.

Wireless is inherently easier to accidentally get disrupted. When wired internet fails, it is rarely the wire. More importantly, if it is the wire that is easy to check and fix.

With wireless internet, disruption can be caused by many things. That makes diagnosing a problem so much harder. Incidentally it also tends to be less reliable.

Besides that, the massive complexity of the technology stack makes wireless harder to manage, configure, figure out compatibility for, and more prone to software bugs.

Wireless does has an advantage in terms of redundancy. If the nearest tower fails, you can probably also use another one nearby. Not so with wired connections, where the whole signal chain needs to be intact. Even if the whole network fails, you can probably switch SIM and use a different network.

I suppose you don't live rurally...

But they both have trade-offs. Wireless allows many connections at once rather easily at the cost of interference and (potentially) speed and reliability, cable is more expensive up front if infrastructure isn't already there.

You can have such redundancy for wire too. People simply figured out that it’s so reliable there is no need for extra redundancy.

Wireless should be the exception, not the norm. When you really need a device to move.

Well I could make that assumption... and make the whole conversation pointless.

I would hate to have all the devices in my house connect directly to a wireless signal from the ISP, for example. I manage the DNS and firewall to control what goes on. Giving up that control means more ads and more tracking and nothing good for me. Having fiber come into the house has no drawbacks that I can see, so why trade that away for something worse?

In dense permanent environments a bunch of cables will most likely be cheaper than having enough 5G cells to give everyone few hundred Mbits of transfer.

In scattered environment the range is too short to excuse the cost of covering it with 5G

There is no niche for it except for "my landlord didn't install proper internet so I use 5G" coz it happened to be alreadhy in range

I could imagine some place like new york, where everyone rents and you are dependent on landlords for your internet contract as well as your actual ethernet connection, 5g for home network could work.

Why would a renter be dependent on the landlord for Internet? I rent (non-US) and have my own Internet, electricity, and gas contracts.

The utilities need owner approval to initially run the lines into the building and into every unit.

You're welcome to get your own contract, but if they don't have permission by the owner to dig into the ground and drill into walls to run the lines to you and a place to put meters and equipment, you're not getting the internet/gas.

Sure, but the gp clearly didn't mean that as they specifically mentioned the contract and the connection separately:

> everyone rents and you are dependent on landlords for your internet contract as well as your actual ethernet connection

Unlike in sane countries where the government has either an independent provider (like Australia's NBN) or mandates the previous monopoly to open up their lines to competition (like UK's BT's Openreach or Japan's NTT), there is no equivalent concept in the US. For calls, sure, there is a concept of an incumbent local exchange carrier, but there's no such thing as an ILEC for internet services.

So if you have a DSL connection, it's tied to a single provider that owns the cable? That would indeed make independent contract quite pointless as changing providers would mean adding new cables.

We have mandated infrastructure sharing both for fixed and wireless networks so the competition is pretty good with virtual operators almost always being cheaper than the brand name owning the infra.

> That would indeed make independent contract quite pointless as changing providers would mean adding new cables.

... if you even have the choice. Many Americans are only served by a single provider, and this isn't the "too rural" excuse but also with many metropolitan areas. There are exceptions, such as with Utopia Fiber (https://www.utopiafiber.com/) in metropolitan parts of Utah, but this isn't the case even in California, New York or Texas.

That's interesting, this is largely routine on the west coast, at least for residential renters. All without owner permission.

I mean you aren’t wrong. But cable has a monopoly on most peoples home internet today. If you could (theoretically) kick their ass with fast internet that completely bypasses asshole cable companies, more power to you.

If you hate the cable company, try the phone company. The cable stops 2 miles down the road from where I live so I have ADSL from Frontier. It was terrible for a long time, I didn’t do much better than 1 Mbps until they added some analog hardware to the line that got it close to 3.

Recently they put in ‘fiber to the node’ to a nearby crossroads, I now have two 18 Mbps lines which, fortunately, give about 100% of the rated performance so I can’t complain much. I can even use Docker now, YouTube, Netflix, and online gaming are all pretty good. (The Samsung TV, outside of game mode, and some monitors, like the monitor on an Alienware ‘gaming’ laptop, add more lag than the DSL does.). I have a load balancer that sends different connections on different lines so collectively the system meets the FCC broadband definition. Still the people down the road get better speeds without the load balancer for less money.

I’d say it is not just a ‘lack of competition’ but ‘pernicious competition’ that is going on here. For instance, the cable company has less reason to serve my valley because people who want TV can get it from satellite. There are plenty of old and poor people who would rather spend $65 a month for 1 Mbps DSL than $70 for 300 Mbps cable and they will go to public meetings to try to keep the better service away. So in some sense competition from legacy satellite and legacy telco makes service worse.

(TV of course is a great case study of ‘pernicious competition’ because of the structure of the industry. Your local OTA TV station tries to keep a secret that you can tune in free with an antenna because they get cash when you subscribe to cable. There is cable, multiple satellite providers, plus a number of IPTV providers but they all provide the same bundle at the same price. There is no way I can say I want CNBC and Bloomberg and the BBC but CNN, Fox and MSNBC can all go to hell. Given that they get paid no matter if people watch they have no need to maintain quality or a brand promise. CNN tried to go direct-to-consumer but decided it didn’t want to threaten the current cash cow even though it may be a doomed business…. It’s hard to tell because there are powerful forces that will try to rebuild the bundle out of Netflix, Hulu, Disney Plus and all that.)

Why would I want to replace my home internet with something less secure, less reliable, lossier, and slower?

The same arguments apply to wireless headphones. (And I'm not disagreeing with you)

No, that's specious. The user experience matters a lot more when using headphones. The weight and tension and tethered nature of a wire degrade the user experience when it comes to an apparatus that goes on your head.

In contrast, I personally do not derive any pleasure from liberating my home internet gateway from a cable that goes to the wall.

> The user experience matters a lot more when using headphones.

Sure, for the use cases where, for example, you're moving too much, wireless is good

But you get an inferior experience on most other cases, for example, switching it from one device to another, not to mention spectrum crowding etc

> But you get an inferior experience on most other cases, for example, switching it from one device to another

I almost always have a better experience switching from one device to another. It’s zero extra taps (just play/pause of some content on the second device) on my Jabra earphones or zero to one taps for my AirPods (one if they’re already playing media on the original device).

Compare that to wired headphones, which I would need to unplug and then snake around my desk to the second device.

That is, if your device supports it, but upgrades take time and my headphones still work great (and weren't too cheap)

Well, the use case where your headphones are connected to one device and you are moving around accounts for, say, 80% of the time spent using headphones. So even if other use cases get more annoying, it really doesn't matter unless they become unbearable (and disconnecting BT on my phone and then connecting from my laptop to the headphones, which is about the worse case scenario, is not unbearable).

I venture to say that in my social circle, over 50% of people own wireless noise canceling headphones.

Yet many of the same people demand wired internet. Hence I don't believe wireless internet and headphones are comparable.

In London, in an apartment block, I tried BT broadband (60Mbps max), then Virgin cable (100Mbps for £55 / month, prices go up for more speed), then EE 4G (£30 / month for 200Mbps +/-50Mbps, unlimited d/l upload is also over 60Mbps, about 6x on Virgin).

I don't game, so don't care about ping.

So, where I am, using 4G for my home internet is faster and cheaper, and I can gift data to my phone. I have not had any issues with reliability.

In Southampton, there are lots of places around the city where 4G coverage is simply not reliable. Even with a strong signal you can get speeds that vary from literally nothing to 60MBPs. I once lived within half a mile of the centre and could never connect to anything between 5pm and 11pm which rendered the service entirely useless.

We've had 4G for ten years or more now, I'm not ready to put my faith in 5G and don't expect to be any time soon.

I almost signed up for a new office a mile from the centre where the best cable speed on offer was 20Mbps and the thought of relying upon mobile internet was enough to make me think twice.

I'm currently in semi-rural location, no mobile service at all indoors but, 200 metres from the fibre cable cabinet, enjoying cheap and reliable 60Mbps without a hitch.

Mobile signal remains spotty in Southampton, I’m unable to use 5G at home because I’m in the dead zone between the coverage of three masts. That’s less of an issue in recent years though because the terribly named “toob” have rolled out fibre across most of the city, and will sell me a line doing a gigabit up and down for £25 a month, which is less than I was previously paying for DSL over a bit of wet string.

Probably council rejecting new masts if it is like most places in the UK.

3UK have thousands (maybe 10k?) new masts in planning to really improve 5G coverage (and they have the spectrum to actually deliver 2gig/sec to phones in the real world). Annoyingly the (vast in many cases) majority will are being rejected on suprious grounds by the council.

The reality where I live (North of England) is the opposite: cash-strapped (and often greasy-palmed) councils will approve masts willy-nilly, to the annoyance of locals who find out only after they're erected. Very occasionally this generates any actual organised blowback.

I'm curious, what are the actual objections to mobile masts being erected? They don't make a noise, they're not (despite the conspiracy theorists) in any way dangerous. All I can think of is that people don't like the way they look, but I don't really see them as being any uglier than a lamppost or street sign.

Mainly aesthetics, the new 5G masts are much larger and need to be sited in densely populated areas where there is already a significant amount of unwelcome street furniture.

Conspiracy theories regarding harmful radiation can also be a factor.

In Copenhagen I tried 500Mb/s cable for about £30 / month, and didn't need to go any further. They since upgraded it to 1000Mb/s.

Several other companies could provide similar speeds at similar (some lower) costs through a fibre connection, which was installed throughout the building.

5G home broadband seems like a fix for when the market for wired options is broken.

I'd be really surprised if you don't have at least one altnet in an apartment building, Hyperoptic is pretty common in London.

I've got access to 4 seperate FTTH providers now in my apartment (all using different infrastructure, not including resellers): VM, Openreach (BT) FTTP, Hyperoptic and now Community fibre.

The limiting factor on speed now is WiFi, which cannot manage 1gig even closer to the router.

It may be worse for the end user but it's great for corporations. Samsung could sell you a smart TV with built-in 5G so that it can always serve you ads even if you never connect it to your wi-fi.

Can't they already do this now? Is 5G necessary?

Well, they can just stuff a 5G NB-IoT modem inside of newer TVs, using eSIMs received by OTA (either WiFi or a previous valid eSIM).

As long as the manufacturer foots the eSIM bill (and it's profitable) they can serve ads.

Or better yet, serve you ads when you're not using the TV where it randomly comes on from time to time and displays an image ad and then turns off again.

Yes, and not just TVs but any piece of electronic equipment with 5G built-in will be capable of serving ads and sending usage information back without needing access to the user's WiFi or router

You shouldn't!

But mature 5g isn't supposed to be any of those.

> less secure

Not really. The link layer is encrypted, and you're almost certainly running SSL on top of it.

Because it's cheaper and more convenient for you ISP and you'll just need to eat the crappier service when noone will want to deploy fiber to your house anymore.

Speak for yourself...my parents live in a rural area and just switched from a local wireless isp (via some directional/long range 802.11 setup) to a tmobile 5g router and it's been an absolute godsend in terms of speed and reliability. Growing up there was depressing and I was flabbergasted to see them join this century with something I'd consider "high-speed" (seen it hit 200mbps personally).

Several of my teenage years involved a lot of conniving to get better internet to our rural house. Best I managed before moving out was very high latency and unreliable microwave at 1Mbps. Now there's great LTE and I think even 5G there.

I’ve been using 5G while waiting for fibre to be laid and use it for day-to-day internet connectivity

It’s 20-30x faster than my VDSL2 line and only £8/month more expensive

And apart from the day of the Queens funeral it’s been really reliable

On the day of the funeral throughput was like 28kbps, whereas 4G from the same tower was 100Mbps so I presume the teleco had an internal network issue with their 5G backhaul / network (probably sold all the capacity to broadcasters)

So whoever doesn't live in the city center gets a poor signal?

Sounds like short sighted view of the decade

Don't be silly. If anyone wanted to live outside cities, people would build cities there.

Before 5g city centers often couldn't cram in enough cell towers to give everyone enough bandwidth for their home connection.

5G has an option for higher frequency towers with signals that drop off much quicker, allowing you to pack towers closer together. Hence giving city centers much more bandwidth.

Down in the suburbs you can still get away with fewer towers with larger range and still give everyone bandwidth.

Out in the countryside is where the problem you propose starts occurring. When the people are so spread out that a single cell is going to be woefully underutilized, and hence not built close enough to your home. But that cross-over is beyond the suburbs, not around the city center.

In getting my home internet from a 4G tower some 3 km away. It's anything between 10 to 40 Mb/s upload and download. I think that we'll never have dense 5G towers here because the area is not dense. At most one tower where most of the houses are.

Supposedly lots of India will never build out data networks that go to homes because it makes no sense when you have LTE and 5G.

Wireless is probably good enough. Wired is still much faster, more reliable, and has lower latency, but if the cost is unaffordable, none of that is essential.

I expect the future is that rural moves to mostly wireless and inner city areas continue to be connected to fiber.

The trend I've seen is the opposite: it's much easier to dig trenches and install cables in suburban/rural locations than it is in a densely packed city, so they often have better wired internet.

It's still more expensive because you are digging 100km of easy trenches to connect a thousand people vs 1km of complex trench to connect tens of thousands.

In Australia we have mostly failed to provide good wired internet to rural areas so they are moving to starlink while the cities are largely connected to fiber now.

There are two kinds of 5G.

One of them is the mm wave which is very high performance over line of site, makes the most of diffraction so it often does well near-line-of-site, then quits working. It is particularly unsuited for handsets because your hand and other body parts can block the antenna. This is a great answer for FWA in favorable geography.

The other 5G is similar to 4G but a bit better performing and my understanding it is actually a bit cheaper for the carrier. 4G infrastructure manages a higher level of spectral efficiency than people thought possible before because of the very close coordination between towers, they actually pipe the baseband signal (around 100 MHz bandwidth) by fiber optic to a central location and process the signals collectively. My understanding is that 5G has a system for coordinating the towers that is still efficient, but not so crazy expensive. So it is ironic that carriers market 5G as something that costs consumers more money because the real driver for it, from the telco perspective, is lower costs.

Worked in Telco too.

I always wondered if it wasn’t enough to have 1 infrastructure provider like with streets or anything else. Why build the infrastructure twice or three times?

Wash your mouth out with soap. I live in Baltimore - if cell towers here got fixed at the same rate our roads do, we'd be an unofficial annex of the National Radio Quiet Zone.

Two wrongs don't make a right..

We used to have that - Bell Telephone. It never improved; there was zero need to innovate. Consumers had to fight like hell for the right just to buy their own phones. Telephones were OWNED by Ma Bell and leased to consumers; their argument was “how can we guarantee stability of the network if anybody can connect anything they want to it?”

[ed] Government broke them up in the early 80’s; overnight we saw innovation like slimline touch tone phones and non-acoustic modems.

You do NOT want to return to that. And the only reason that you like monopolies on utilities is because it’s been like that for generations; we honestly don’t know what innovations would happen if utilities and roads were privatized and not enforced by government monopolies.

> You do NOT want to return to that.

I'm not sure your assessment is reasonable. You're putting up a strawman by claiming that the only way to have shared infrastructure is to have an abusive monopoly that stifles innovation and screws consumers over. This is false. It's quite possible to have a single infrastructure provider that does it's job, such as a consortium of telcos shipping in to deploy and run shared infrastructure. There are plenty of examples of those types of initiatives in other areas, such as shared ATM networks, which provide a far better service than "everyone for themselves" deployments.

But just imagine. All of the incumbent providers use shared 1gb fiber. That’s what you get. You can pick your provider, but it’s all the same fiber infrastructure.

Some scrappy company comes in saying they can offer 10gb fiber for the same price or less, but it takes an infrastructure upgrade. The consortium of incumbents shoot down the idea because it will “cause problems” or whatever. So they’re never forced to spend money to upgrade the shared tech, and your service never improves.

It's quite possible to have a single infrastructure provider that does it's job

I don’t doubt it. But do you have any examples of this having ever happened? Every case I’m aware of has gone much like the person your replying to describes.

For example, I moved to England and spent the better part of ten years trying to get British Telecom to install a working telephone and internet connection at a house in the center of a large town. They’d send a truck out a few times a year to fix the same line fault, which consisted of having a guy climb a pole, unplug somebody else’s house from the “good” connection, and plug them into one of the bad ones. Presumably that neighbor would get the truck to come back at some point, starting the cycle again.

There are no other providers. Or rather, the other providers all subcontract from BT, so you eventually get the same truck to come out. But now you’ve added another layer of pain to the equation.

If there were a second option, they would get all the business. But there is not, so you’re stuck with what you’re given.

> But do you have any examples of this having ever happened?

Yes, there are whole Wikipedia lists of this having ever happened.

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

This is no rocket science. I mean, even utilities grids in open markets work like this, for stuff like power and even gas. Nowadays even railway networks in Europe are decoupled from operators. Why do people still cling to old monopoly tales?

Because there is a huge infrastructure dependent on old money continuing. They own the financial, the judicial, the media, and the educational institutions, and they insure no other viable alternatives to the mainstream economy they control is even taught as a viable alternative.

Demand for power and gas—or banking services—doesn’t scale by orders of magnitude every few years.

The FTTH networks in Scandinavia are more than 20 years old now. The UK fiber plans were abandoned what, 30 years ago? Demand for bandwidth grows faster, yes, but it's not like you can't lay a larger pipe to begin with...

“how can we guarantee stability of the network if anybody can connect anything they want to it?”

Sounds hauntingly familiar.. "How can we guarantee the security of the phone if owners can just side load anything onto it."

https://www.theverge.com/2021/6/23/22546771/apple-side-loadi...

Singapore does a decent job of it. ISPs provide service over shared fiber infra - https://www.netlinktrust.com/

They don't have to be at the cutting edge of future innovation. If customers in the USA or somewhere else, who are paying much more for their broadband, start benefiting from some new technology that requires changes to the shared infrastructure, then Singapore can implement it at that time.

It's not as bad as you think. A lot of towers are owned by someone other than the carriers, and the all lease space on the same tower.

Luckily it's not all wasted effort. The different networks boil down to implementing a very simple FDMA scheme. Each telco gets their own frequency, and customers of different telcos never interfere with eachother.

I hope/imagine that a lot of fiber backhaul is shared between operators that share towers.

> I assert that it is dawning on operators that consumers are not interested in paying $10 extra per month for 5G

I agree with this. For the most part, LTE is enough for everything from casual web browsing to streaming video. There just isn't that much of an improvement jumping to 5G in the end-user experience. The is very different from the 3G to 4G jump. I still remember using an original 2G iPhone. Maps basically couldn't load over 2G.

Also, providers by and large still haven’t delivered on the original promises of 4G.

I’m old enough to remember the self-driving cars and service robots that UMTS (=3g) was supposed to deliver

Most new cars these days do download maps, traffic, OTA updates etc using 3G, so yes, this did happen

What were the original promises of 4G?

100Mbps down and 50Mbps up.

In some places they did deliver on it. I understand that cellular data is pretty shitty in the US, but in many places in EU you’re getting 100mbps with unlimited data for like 20€ a month. I could even get 200mbps if I pay extra.

In Canada we call that $1.50/s down, $0.75/s up.

5G is great for those who want > 100MBs internet connections without having to be locking in to a multi-month contract with a fibre broadband provider.

You can just buy a SIM card and you’re on.

If you live in the right place, that might be true.

I have 1G fibre on a contract I can cancel any time with no fee.

I was getting 800mbps on 4G though.

> Verizon have some high bandwidth spectrum, but it doesnt propagate as well as lower (and slower) frequencies like what Tmo has.

Eh.... nitpick. Let's ignore mmwave for a minute and focus on bands that matter for most people. T-Mobile and Verizon both have midband spectrum which is overall most useful. Verizon's is C-band 3.7ghz T-Mobile's BRS 2.5ghz midband is mostly backhaul limited where it's deployed, it'll do ~700mbit/s down but the actual spectrum is capable of at least double this. (in my market I can get 100mhz + 20mhz of N41+N41 aggregated, this is 2.5ghz spectrum). It doesn't matter if I'm next to a tower, it's limited by the backhaul.

I get between 600-700mbps real world if I stand outside my house. I live in semi-rural area and the tower is about 6 miles away nearly line of sight. Indoors the attenuation hits and I mostly use the other bands to reliable pull ~150mbit/s, which is fine for me (and traffic goes mostly over LTE).

> - Telcos desperately need a killer app or use case that drives 5G adoption. They havent got one.

The 'killer app' is called home internet in markets that have trash CLEC's which haven't upgraded infra in 20+ years. Mine is Blightspeed selling 'up to 3mbit/s' copper DSL in 2023. Other option is Comcast, which works but is ludicrously overpriced because there's nobody el> Verizon have some high bandwidth spectrum, but it doesnt propagate as well as lower (and slower) frequencies like what Tmo has.

You have to pay extra for 5g in the US?

Some do, but AFAIK T-Mobile doesn't

Here in Australia 5G is awesome:

https://imgur.com/gallery/BFTTMRS

Measurement was taken a decent distance from the city in the suburbs.

Whoa what service are you using? I’m on Vodafone 300gb plan and it’s always one bar on 5g.

I rarely see speeds that fast too; downloading a game pass game is usually 1MB/s for me

This is a story I see a lot in the US. Two customers, even with the same carrier, seem to get very different coverage. I thought circa 2010 that I got the same coverage with a prepaid as a postpaid phone but starting around 2018 I would go to New York City and have trouble w/ my prepaid phone.

There seems to be some phone and plan that gets better coverage than another but it is by no means documented. Maybe it just ‘works better the more money you spend’ but without any explanation of what the rules are I am afraid of spending a lot and still getting bad coverage.

Telstra post paid, 100gb/month for $70 AUD/month

> This is a bit of a problem when those same operators spent or borrowed billions

Not to worry, they'll just pass the cost to customers or ask for a bailout and get it. Being a giant corporation is awesome.

In London 5G is definitely way better, tested with Vodafone and Virgin (O2? Voda?)

Here in the US I think the killer app is breaking the local cable company’s monopoly on providing high speed internet to the home.

Somewhat, but the carriers are kinda half-hearted about it, particularly when they talk with investors. FWA users use a lot of gigabytes per dollar that they spend and if the carrier thinks in terms of that metric it may see it as bad for business, possibly cannibalizing more expensive mobile plans.

FWA is a spotty solution anyway. Economically it works best in the ‘WebPass model’ where you server a whole apartment building in an urban area that has a clear line of site.

Rural carriers are averse to maintaining their copper networks, never mind installing fiber. If you are fortunate you have a line of site to a tower but probably you are less fortunate. When you head south from Ithaca, NY you encounter a dissected plateau at an altitude around 2000 feet. There is no ‘highest hill’. A tower on one ridge can sorta see into the immediately adjacent valleys (though the sides may be steep enough to still make shadows) and under 1 GHz maybe into the next valleys over. Heading South down to Alabama the mountains get bigger but it is basically the same story, where FWA goes to die. It is delusional that it could be cost-effective or even effective universally because in a lot of cases there isn’t even a road or a power line at the top of the ridge.

FWA really does create competition for people where geography is favorable but it is no answer for the many people still stuck with slow DSL.

> Rural carriers are averse to maintaining their copper networks, never mind installing fiber.

My rural area is served by an independent telco/ISP. They would agree with the part about maintaining copper, but installing fiber has actually been their solution due to the lower cost of ownership over the long term. I wish more carriers saw it that way.

I went to tmobile's 5g internet after Comcast refused to negotiate better terms on renewal. I live in a HCOL metro but have yet to be hooked up to the local fiber company, so although this 5g home internet isn't so great (high latency, trash-tier upload speeds), download is good and it is great to have an option against the Comcastopoly.

Except what's the point? You're just ditching one terrible communications megacorp for another.

I live in a rural area, so I obviously have my bias associated with that. But for me, investment in more coverage would have much more value than investment in higher bandwidth for areas that are already covered.

Either way, my 5G iPhone is far less capable of connectivity out in the world than my last one, more so if I go indoors. Even just inside a door.

I live in a rural area (not a single Starbucks in the county), and in 50% of the town, my phone has 5GUW coverage. I can download apps, playlists, and movies quickly. I recently downloaded UPBGE (430MB) on my iPhone in only a few minutes. This wasn’t possible on LTE and is worth the extra $10 for me.

I find your definition of a “rural area” hilarious.

The lack of a corporate coffee store does not necessarily mean it’s a rural area. Been to plenty of small population counties and there was at least one starcrack store in the area, including a McDiabetes, and other low cost crap.

Usually a rural area is defined upon the population size and density.

I'm a bit confused about your last point. Could you elaborate?

Telcos are (in theory, at least) private utility companies, so why does it make sense for them to create the application/use case? Do you mean that Telcos like Verizon are reeling from not creating their own platforms (say, apps or app stores) that charge for use of LTE etc.?

Basically, why and how do you think that Telcos are reeling from not capturing LTE value? Didn't they already by charging for use of their network?

Side-note: If I'm not wrong in my assumption of what you meant, then you conjecture a rather depressing idea that companies feel the need to capture value from an increasing number of market segments and crave hegemonic monopolistic control, for example Microsoft since forever.

Telcos set up LTE at huge cost, drove up prices a bit (as much as they could bear given competition) and the over-the-top content cashed in on having much more bandwidth available.

In their ideal world, they would put up the investment for the infrastructure (or even better: have gov't pay for it) and then turn on the money printer.

Basically, telco is a fee for carriage service. You pay for transport on some network that they build: cellular, fiber, etc. They figure out what to charge you based on historical models of what you used the transport for.

In the days when all that you transmitted were telephone calls, this was easy. Then people started sending IP packets and it got a little harder, but ok, the telcos would just charge a per byte fee. The money was in charging to carry your phone calls and SMSes anyway.

Then the smartphone exploded and suddenly consumers did not give a hoot about phone calls. The sole reason to have a cellphone plan was to get data. The telcos had to expand the network capacity by several orders of magnitude for a customer base that had absolutely no desire to pay by the byte. This blew up their model. Moreover, telcos saw that people were paying $$ every month to folks like netflix, watching video on their phones, chewing up network bandwidth, and the telcos werent getting any piece of that action. Hence the whole net neutrality thing.

The aim isnt that telcos have app stores. They know they cant compete with that. The holy grail is that a telco finds some use of the network that commands a premium price that users will accept. Verizon trades on having the best-est network. AT&T tries to bundle, eg HBO. Tmo goes after the younger customers with discounts.

Now 5G is here and telcos are desperately, desperately trying to find uses for it that cant be satisified by the existing network and biz models. Will they succeed? I wish I knew.

Ah, thanks for explaining that. Really appreciate it. Makes a lot more sense now.

OK, I will bite.

Comparing 5G vs 4G as Technology is completely different to 5G vs 4G in real world implementation. That is a bit like asking if Core i9 is faster than Core i7 without knowing the clock speed, core count, cache and power budget. Or for Software developers comparing Python vs Ruby without knowing the code and VM, JIT or not

Most 5G network around the world are deployed in higher frequency spectrum. Hence the receiving Bar ( signal quality ) will generally be lowered until they are deployed in sub 1Ghz Band. Refarming frequency from 4G ( or 2G and 3G ) to 5G takes time, MNOs using Ericsson could use Dynamic Spectrum Sharing ( 4G and 5G inside the same Spectrum ) for faster 5G Rollout. But it has its own set of problems.

Until more users switch to 5G capable smartphone, and MNOs work their way to switch to 5G and especially 5G NR SA ( Stand Alone 5G without relaying on 4G Network Back End ). The full potential and advantage of 5G won't be noticeable if not, as in your case even worst than 4G.

Before anyone ask why switch to 5G then if it isn't better now. Well it really is a Chicken and Egg problem. But one way or another MNOs would like ( force ) you to switch to 5G as 5G offers better Network cost efficiency and much higher capacity.

> Stand Alone 5G without relaying on 4G Network Back End

What's the difference in the backends? Is it just a question of bandwidth, or is there something more?

Different signalling, better privacy/security (resistance to IMSI catchers), lower latency. It's not a ton better in bandwidth for sub-6ghz but the latency improvement is noticeable.

I still use 3G intentionally until it is sunset or disconnected. It uses less battery power than LTE, and comes with the nice advantage of that it doesn't support alerts. The latter is an issue because in Canada, the absolute morons who made the decision, decided that all alerts, every last one, are to be sent at the level of Presidential.

So, this means an amber alert for some kid who is with one of their parents 500+ miles away blasts through.

So people are forced to do complicated stuff like rip out the alert module through the terminal, or set up some sort of schedule where the phone is airplane mode or turned off at night, etc.

Too much hassle, 3G is easier.

And 3G does use less power. LTE is faster with lower latency but beyond that.. I don't have VoLTE either because the provider is being cheap with my cheap plan.. so why bother.

5G will have lots of battery consumption if the example of LTE is any indication, and no user-facing improvements that I am aware of. Real ones, not theoretical. Wireless data prices are very high here. Nobody is going to use 5G for an internet connection.

The difference is quite large. The architecture for the 5g backend was made to allow for easy virtualization and scaling capacity up and down. Hence telcos don't need to buy a very specific subscriber management box that is special made and needs to handle peak usage. Instead you can buy or rent commodity hardware, and usually run much less hardware because you can scale up as required.

Check out the diagram in this article: https://www.geeksforgeeks.org/5g-network-architecture/ Note how little space is given to the cell towers.

I just don’t feel like there is much need for faster than 4G right now. It almost feels like a meme to say this “why would anyone want more than 16 mb of ram”. But it’s been years now and not much has changed. I did a speed test and my 4G was 300mbps, that’s faster than the home internet speed of most people I know. I just can’t see why you’d want more than that on a phone.

Yeah, customer does not care where is a problem. Customer sees 5G on his phone, customer can feel that it is not faster than 4G (because it is that 5G on Sub 6GHz), customer has negative feelings about 5G because he does not understand why he should be paying more for same speed.

I'm not sure there's anything wrong with that. Technology has always been only as good as its implementation.

> Or for Software developers comparing Python vs Ruby without knowing the code and VM, JIT or not

As far as consumers are concerned we are comparing Python 2 and Python 3

You're saying 4g was actually better in every way, but we'll be forced to give up and accept 5g?

Does 5G at least help with the push notifications? I remember reading somewhere that one of the reasons why non-centralized push notifications suck was that the telco hardware kills the connection (presumably due to limited memory)?

In my day to day, 5G uses more battery, has similar speeds to LTE, and is less reliable than LTE... I just go into settings and turn off 5G.

Occasionally I'm at the airport (or someplace with the "real" wideband 5G) and I want to download a movie quickly. At that point I turn on 5G and get 2 Gbit down, which is convenient!

Is the only good usage of 5G to download movie on the way?

I don’t know many other things that would need 2gbs on a phone.

A dirty movie on go

I use my smartphone on 5g (with unlimited volume) as a modem for my laptop when I'm not home. It's fantastic. I'm glad nobody here is in a position to decide "nah, 4g is good enough". Nothing but luddites here.

The actual real benefit of 5G for me is beating very congested LTE towers at peak times. There are a lot of spots where the 4G mast (especially on LTE800) is completely congested, but 5G is much* better, though definitely starting to see 5G congestion in places now.

In time as people upgrade their phones the balance should shift then, with the 4G spectrum becoming less congested. Personally I don't find the speed increase of 5G to be a need as I never found 4G to be slow. I'll be quite content to lock to the 4G bands if they open up more.

No, because 5G at the moment combines both 4G+5G carriers together. This also seems to reduce performance of 5G on heavily loaded 4G cells (probably ACK packets getting delayed if it goes via the 4G upstream is my guess, but not sure).

It wouldn't be so bad if phones were better at switching to the best available signal, they they could use the efficient 5G connection when it's available, but fall back to 4G when the 5G is blocked by a building or something (which I gather is more common due to the high frequencies used in 5G). But they seem to insist on connecting to a crappy 5G signal even if an excellent 4G signal is available. Same with 3G, which is often fine for many tasks.

This is inexcusable. I thought my phone was defective for weeks. I would get ZERO data, move 100 yards and have data again. After some Deep Thought™ I disabled 5G and could magically use the Internet from my phone again.

One of the things that sucks are devices that use standalone 5G but don't support carrier aggregation on those bands. It's possible network will steer towards an over-congested low-band or a weak mid-band signal, but it cannot combine and make the best of both. These devices are better served using non-standalone 5G or sticking to LTE.

Same experience here. It was extremely frustrating to realize that I have to play a game of whack-a-mole with my phone's settings to make the damned thing usable.

I have an iPhone 13 Pro Max here in Taiwan. If my 5g goes to 4 bars it switches to 4g.

Is that what you mean?

Yes, please share your secrets with us!

I wonder if 4G is better now that people who are really enthusiastic about bandwidth usage have moved up to 5G.

For most operators, spectrum is dynamically shared between LTE and 5G, so probably no.

Most? Verizon uses it heavily, AT&T uses it slightly in a few of their markets but has been phasing it out, T-Mobile does not use it.

Phones seem to optimize for the fastest network instead of network quality. At least on Android it's possible to manually override and select a network through the hidden settings menu accessible through the phone dialler. Just hit ##4636## and you're in.

Can you explain more provide a provide a link. This is the bane of my existence. My phone has two towers in range, one with essentially no signal and another one with four bars and it prefers the low signal Tower. It also prefers my home Wi-Fi even if I am taking a walk around the block. It is infuriating that I have a 50/50 chance of resolving a website like Hacker News with a flagship smartphone in the Bay Area

HN is eating asterisks in the code. See https://www.todoandroid.es/en/android-mobile-code-what-is-it...

Yeah I have this issue with WiFi as well, every time I'm leaving my house I pull out my phone as I'm walking down the street to play music but I have to wait for the one bar of WiFi to disconnect

Does not seem to work on Samsung and Huawei.

Unfortunately "switching to best available signal" is a really hard problem. More so for a device in motion and on really short waves like Ghz frequencies.

I always get annoyed when people nag around about WIFI - because they cannot understand how many things happen in radio communication even if you sit behind the laptop and you have router in other room and you have 10 neighbors having their laptops/routers also running. They expect stuff should "just work" and reality unfortunately is much more complicated.

> I always get annoyed when people nag around about WIFI - because they cannot understand how many things happen in radio communication even if you sit behind the laptop and you have router in other room and you have 10 neighbors having their laptops/routers also running. They expect stuff should "just work" and reality unfortunately is much more complicated.

Oops, maybe don't have a look at my comment about then then[0]. :)

[0] https://news.ycombinator.com/item?id=34569334

I thought that might be the case, but if higher frequency 5G signals are susceptible to small black spots due to buildings etc what's the solution? Presumably they aren't going to deploy enough 'microcells' to cover every place that the cell towers can't reach, so are we stuck with more dead zones if we use 5G?

Side note: I have the same rant for laptops, wifi and 5ghz/2.4Ghz. If a network is on both most clients will just prefer the 5ghz no matter the signal quality/packet loss.

As a user I'd rather have a working slightly slower connection than a broken faster* one.

*Not actually faster due to packet retransmission time from packet loss

Even on 2.4GHz the internet connection is usually the limiting factor anyway, unless you're on a high speed network with a lot of congestion.

So this is basically boils down to false advertising, which is the real problem here

One of the reasons 5G seems worse is because most implementations are NSA (non-standalone) right now. As others have mentioned, this means you're using the LTE network for control/signaling and 5G just for data traffic. 5G SA improves latency because it ditches the LTE signaling but it's not available everywhere on most operators yet. Verizon and AT&T are just starting to implement it. T-Mobile US has been ahead of the curve in offering 5G SA nationally for a bit now.

However, probably the most important reason 5G performance can appear worse is because NSA works via a technology called ENDC - EUTRAN New Radio Dual Connectivity. This enables your device to receive LTE from one cell site and 5G from another. LTE being there to carry the control/signaling as previously mentioned as well as improve speeds by combining LTE and 5G channels.

When you're receiving 5G from one site and LTE from another. Depending on your device and your network operators settings, more of that traffic will flow over either LTE or 5G. If your 5G channel conditions are poor (say 5G is coming from a far away site but you're also on a closer, LTE-only site), but your device is still preferring to send data mainly over the 5G leg of the connection, you could potentially see worse performance than you would on LTE alone.

LTE Advanced (LTE+) is the sweet spot, they have nearly all the bells and whistles of 5G like CA, MIMO, etc. 5G is over-engineered, it's more like LTE advanced with new frequency, beamforming and some new channel coding while consuming much, much more power both on your cell-phone and cell-towers. The speed and latency benifits are almost negligible because the metrics advertised by telecom industry are measured only between the SA 5G devices and 5G towers, in reality the latency and bandwidth are deciede by the hops all the way to the IDC (unless the so-called "edge-computing" is a thing yet they aren't), so the 5G promises may never happen.

This combined US/EU boyccott of questionable but cheap suppliers like Huawei makes the cost of 5G not worth it.

It's not strictly worse than LTE, but it's not quite just you. A lot of it will depend on a bunch of factors.

If you're in an area with mid-band 5G (5G UC on T-Mobile USA, 5G UW on Verizon, 5G+ on AT&T), 5G will provide a meaningfully better experience most of the time. TDD (time-divided rather than frequency divided) 5G on mid-band spectrum is seeing great results. T-Mobile's 5G network is averaging 217Mbps (including both their low-band and mid-band networks, but it shows how much mid-band 5G can do).

If you're in an area with low-band 5G, it's complicated. If you're on Verizon, they're using DSS (dynamic spectrum sharing) to use the same spectrum for both LTE and 5G. This means that the network is constantly context switching between LTE and 5G and it provides pretty crappy service. If you're on T-Mobile, their low-band 5G network is using new spectrum and does often offer some moderate advantages. Another complication is that you might be sharing that 5G with all the heavier users while an LTE channel is less used. You could also be connecting to a tower farther away to get a 5G signal.

It's probably not that it's higher frequencies since you're probably in an area where it's using similar frequencies. I'd say that this is more likely carrier-related than 5G related.

Depends which flavor of 5G you are talking about. mmWave is still terrible unless you are right next to a cell. Midband is slower, but a huge improvement over LTE, and uses adjacent bands in many cases.

Verizon and ATT pursued a mmWave first buildout, and now they are catching up on midband. That’s why the C-band issue was so huge last year for them. Unique in the US, T-Mobile did midband-first (it was the point of the Sprint merger —- to get Sprint’s frequency licenses that were good for midband), and they have the most useful 5G network, mostly because of availability.

My iPhone will say “5GE” in some areas.

My iPhone is physically incapable of using 5G. It has no 5G antenna. AT&T just makes it say 5G anyway. They use this to “advertise that 5G is available in that area” or some other weird lie. It’s not 5G for me, and I honestly doubt it’s 5G for others either

Completely anecdotally, the internet in these “5GE” areas is usually worse. In fact, nearly everywhere I go anymore my cell service is atrocious now. I have to actually seek out WiFi networks to get usable speeds sometimes. My cellular speeds have noticeably tanked in the past 2 years or so. I used to get like 50mbps sometimes when doing a speed test. Now I usually get ~1mbps at restaurants, or less.

Same. I get 1.2Mbps with ATT on 5GE in Seattle metro.

Its been a clear upgrade for me. I am on T-Mobile and in my apartment i see around 200-300mbps and across Seattle metro area i hit 200mbps consistently.

T-mobile has better coverage here compared to ATT and Verizon because of Mid-band spectrum they got with sprint merger. I think other 2 are in process deploying more mid band spectrum and may be that will improve it in future.

In Seattle I see 5G UC (Marketing term for their MidBand Spectrum) for TMobile pretty much everywhere.

I use 5G home broadband because I live in an old building with rusty DOCSIS 2 cables (and also because there was a risk of being sent to quarantine camp in recent years). I get a very stable 200mbit connection and can ping Google with 5-7ms latency.

Note that I live in one of the most busy parts of Hong Kong so YMMV elsewhere. In case anyone's interested, I'm using SmarTone which is the only provider in HK offering unlimited full speed 5G

Depends where you are, for me in the center of Hong Kong, with a Huawei phone, it's a net improvement, it's way way faster. I even use it instead of wifi on my phone because my router slows down with distance while 5G stays fast in every room.

It consumes more battery so when I need to save I go back to 4G and I notice the slog.

Maybe your phone doesnt have a fast chip, your city isn't dense enough to have budget for antennas, or the providers doesnt want to give you 100Mbps... I was telling a friend recently how much of a miracle the internet speed on phones were these days, to tell you how 5G is positive here. It's not even that expensive.

In case it's relative and you expected much more, here is a speedtest from my phone, middle of Central HK, 40th floor inside my toilets, where I often browse HN from: https://www.speedtest.net/my-result/a/9021015738

In center of Taipei 5G is spotty, but go out to a country side city and I’ve got 5G everywhere. Go figure :)

I have a Pixel 6A on T-Mobile. I keep 5G turned off because it's usually slower and flakier than LTE, visibly chews through the battery (a percentage point every couple of minutes), and heats up the phone.

Are you running the latest Android release on it? I had a Pixel 6 Pro on T-Mo and experienced almost identical issues until I updated it some months ago

Turns out Google pushed the 6 series out before the OS was "ready" and "forgot" to tell anyone

Anyone know if there's a way to disable 5G on a Pixel 7? Using Visible, a Verizon MVNO.

For me, I switched the preferred network type to LTE in Settings > Network & Internet > SIM > Preferred network type.

Most providers have optimized their 5G service for dense urban environments where you have direct line-of-sight (meaning no walls or even foliage) to a nearby transmitter. The idea is to improve the experience in major cities and at event centers. T-Mobile is the only provider I'm aware of that has optimized for higher performance when you're not next to a tower. They operate a long-range 5G network that is a direct upgrade from their long-range LTE service.

tmobile's also the only one with a low frequency (600 Mhz vs Ghz) 5G spectrum to do deployment in/under.

It can be a factor of infrastructure as well.

When 4G first started I noticed really fast 4G service at 2am in the morning, but super slow during the day.

Most of the cells I was connecting to were in the city.

Over time as they built out the 4G infrastructure it got better, but for well over a year if you were at work and struggling to get mobile bandwidth then a switch to 3G was almost always significantly faster, double upside because everyone wanted to be on 4G and deserted the 3G bands, so no bottlenecks.

I feel it heavily depends on the modem. I used to have a Pixel 5a, which has an SoC (Snapdragon 765G) that was released during the late stages of the transition from LTE to 5G. On that phone, 5G was a terrible experience and my data wouldn't work half the time when I was outside. I switched to a Galaxy Z Flip 4 a few months ago, and 5G is just fine for me since then (i.e. I don't notice a difference compared to LTE). Same carrier, same locations.

I wish I could get a 2G phone. 3G I'd settle for. All I want is email, Maps, Uber and a Camera. Everything else is just addicting time wasting behavior. I love when my plan runs out of fast data and puts me on that throttled level.

Light Phone exists, don't know how good it is.

But email, map or uber use tons of data where 2G and 3G were not decided for? So you will still need a 4G, no?

email: only for attachments, not for text

map: preloaded maps were a thing before. So, no.

uber: preloaded maps. No, again.

In a lot of places there are no longer 2G networks.

Android 13 has eBPF-based network throttling options in the developer options, which will give you a similar 128kbps experience without sacrificing battery life.

This applies on all interfaces, including WLAN, with the affordance of being able to switch back without dropping flows.

I wouldn't recommend it for WLAN from a total system capacity perspective, as throttling eats up air-time.

The point is I have no self control. If I could manually switch it I can't be trusted.

Uninstalling the apps helps somewhat. Facebook and Reddit have terrible mobile web experiences, which helps even more!

This was true until I discovered mbasic Facebook :(

Oh you can get one except us providers have turned off their 2g base stations

It's 3G that they've disabled, 2G/Edge still works and they're probably not going to actively tear it down since I recall it e doesn't use up any additional bandwidth (something about how 2G exists in the guard bands). That said I doubt they're going to pay to replace the infrastructure for it as it begins to fail.

Nope, not just 3g. Both verizon and tmobile shut down their cdmaOne stations (along with cdma2000) last year and att I think a while before that.

Lol, 2G for (2023) Uber? I don’t think you know what you’re talking about. I was once rate limited to 2G speeds due to a problem with the carrier, and Uber took like ten minutes to get past a blank screen.

Fantastic here, but heavily area dependent, low-latency and I often see speeds above 250Mbps.

20Mbps is not enough for a phone I use for tethering. 5G home internet is a thing here now too (Australia), and in a lot of cases beats the alternative.

Wonder what the difference of electricity cost is between running your home router via 5G and wire.

Outside (and not that far outside) metro areas in Australia, this wire you speak of is unknown.

In places I’m looking to live, your choices are 5G, “Fixed Wireless NBN” which tops out at around 60-80Mbps and Starlink. I know starlink takes a lot of power.

I see some 5G New Radio routers shipping with 12 volt 1 amp power supplies, and others with 12 volt 2 amp power supplies, so 12 to 24 watts at most.

Its impressive that you can get a gigabit, WiFi 6 capable router with a new Qualcomm 5G/LTE chipset to fit in such a small power envelope!

In the UK it seems a clear upgrade. I can be sitting on a train downloading Netflix shows between stops before we go underground. Speed tests often higher than I get at home over VDSL. Couldn't do that on 4G.

> the higher frequencies used for 5G don't go as far or penetrate as effectively.

Not true. 3G & 4G bands are being refarmed to 5G, I am seeing freqs as low as 700 mhz being used around here, so they'll penetrate far and wide.

That said, the presence of a 5G icon on your phone doesn't mean you're actually using 5G, that's very cheeky from operators.

The really meaty 5G band here in UK is N78, 3500mhz, some operators got big chunks of those, for example Three UK has 100mhz of that, contiguous, and people often see 1Gbps+ downloads. I found this frequency does a reasonable job and inrural areas, with high gain antennas, you can still receive it miles away.

mmWave bands - not currently in use in Europe, will indeed be only useful in small (and crowded) places like stadiums, shopping centers, etc. Of course, they could install them on top of many street lights or telegraf poles, then they will become much more useful, in reach of most of us - basically very high speed, low latency "broadband", wireless.

5G is better than 4G in India. Obviously, there are signal issues in remote parts and it isn't available in all the towns. I get 1000Mbps in my home and that to for free at the moment. Thanks to Jio. I have a 200 Mbps fiber connection as well. When I was on LTE I used to switch to WiFi at home but on 5G I don't feel the need to switch.

I was skeptical about the utility of 5G until I downloaded the whole of the Witcher 3 onto my Steam Deck through my phone at 200mbps while on holiday. To be honest I’m still somewhat skeptical, and maybe I should have left the Steam Deck at home instead of taking it on holiday, but I was very aware in that moment that I was living in the future.

Unlimited data? Sounds cool until you use up all of your data.

Are you using T-mobile or another provider that uses them?

My 5G connection is trash, but I just switched up Google Fi, which is using T-Mobile's network.

Inside my apartment pictures on websites are loading like it's the early 2000's again.

I'm in a major city, I thought that would help.

I'm on Visible, which is a Verizon MVNO.

I use both Visible and Fi (= T-mobile) on one phone, I wish there were a way to fuse the networks and switch packets with millisecond-level failovers so I don't have to keep swapping the active SIM.

There are services that offer bonding like speedify but I don't think iPhone is supported for bonding multiple cellular connections. Not sure if Android phones do. Maybe someone has experience with it?

The iPhone isn't using bleeding edge 5G and doesn't let you lock to 5G Standalone, from what I have seen on Android using 5G Standalone (which nixes bonding with LTE) gets rid of most of the jitter that causes data connection lag.

The IMS core that powers calling and texting on most cellular networks has not been upgraded to route calls and texts over 5G Standalone yet (or if so, it doesn't seem to work with the Pixel 6 or 7 on T-Mobile or AT&T), meaning most phones stay on 5G Non-Standalone which is a substandard experience.

> The IMS core that powers calling and texting on most cellular networks has not been upgraded to route calls and texts over 5G Standalone yet (or if so, it doesn't seem to work with the Pixel 6 or 7 on T-Mobile or AT&T), meaning most phones stay on 5G Non-Standalone which is a substandard experience.

This is not true. T-Mobile has been allowing voNR calls on their network for over 2 years now on SA. It's up to to the handset to have it configured in modem firmware (it is configurable on Qualcomm chipsets through EFS). AT&T appears to be allowing voNR calls since around August of last year.

As for the 5G NSA substandard experience, most phones on the market don't support SA carrier aggregation. Without that the experience sucks, This is a hardware feature you'd need a flagship phone built within past 2 years to support (new ones do 3CA, older ones do 2CA). Better to leave on NSA for these phones or stay on LTE until more spectrum refarmed to NR.

Speedify lets you bond the cellular to the wifi. It doesn’t let you use the LTE AND 5g from you phone at same time. But it does let you join another phone’s hotspot so you can combine your cellular with the other phone’s cellular.

What make and model is your phone?

Visible is a deprioritized version of Verizon's cellular network, so when your on a busy tower you will see slower speeds than postpaid Verizon Wireless customers. Verizon also happens to have the slowest network over the last few years since they have the most customers using it (nearly 130 million users) and smaller spectrum holdings than the other two major cellular networks.

Verizon also chose to deploy 5G service using spectrum sharing, so 4G and 5G speeds are very similar: https://www.zdnet.com/article/verizon-5g-dss-isnt-the-5g-you...

T-Mobile is using dedicated (and much larger sized) low and midband spectrum for their 5G connectivity, plus they have almost half as many customers, resulting in much faster speeds.

That being said 5G is still in its infancy, 5G Standalone is really needed to lower your phone's power use (so its not connecting to LTE and 5G at the same time), reduce latency and jitter, and allow 5G to stretch further for a more stable connection.

Thanks for the info.

I'm on a Pixel 7. Honestly I'd like to try disabling 5G but haven't been able to find a way yet.

They added it! It was not possible when the phone was released, but one of the updates (December maybe?) added a toggle button in settings.

Same here, on Visible. The 5g sucks, and I can't even disable it on my S22.

I made custom APN which removes 5g and it's just ignored =(

This is likely your problem.

My first question is whether you're on one of the new plans or one of the old plans. This is actually important because the old plans are provisioned differently. The old plans use Visible's cloud APN while the new plans run off Verizon's. A lot of people have complained that Visible's was pretty terrible. The old plans were $25/mo if you had Party Pay. The new plans start at $30, but don't require any hoops.

After that, I'd note that if you're not on the Visible+ plan, you are going to be on a lower QCI (Quality of Service Class Identifier) than most of the traffic on the network. So if others are using the network, you might experience slowness, lags, and hangs. Verizon's network isn't in the best capacity position in a lot of areas so this can be meaningful.

Verizon's low-band 5G is also not great since it uses dynamic spectrum sharing rather than having dedicated 5G spectrum.

If you're on one of the old Visible plans, I'd pay the extra $5 and get on one of the new plans. You'll need a new SIM card for it, but it's probably worth it to get off the cloud core that has generated a lot of complaints. If you're on the new $30 plan, it might be worth trying the $45 plan for a month and seeing if you get better results with the 50GB of premium data. After that, you should probably try T-Mobile or AT&T.

You can try AT&T for free using their 14-day free trial on Cricket (https://www.cricketwireless.com/free-trial.html, note that it's only 3GB of data) and you can test out T-Mobile for a whole 3 months if you have an eSIM device (https://www.t-mobile.com/offers/free-trial). If you can, it's definitely worth trying an alternative carrier since they're now offering free trials (Verizon also offers one for 30 days: https://www.verizon.com/support/verizon-test-drive-faqs/).

EDIT: as an aside, it's probably not that "higher frequencies used for 5G don't go as far or penetrate as effectively" simply because, unless you're on Visible+, you don't have access to any of those higher frequencies. Legacy Visible plans and the current $30 Visible plan only have access to low-band 5G using the same frequencies as LTE.

Awesome, thanks!

Im a 5G Google Fi user (Pixel 7) in Boston (I live in the city) and its honestly trash. There have been so many times where it says 5G but there's actually no network (maps keeps using the downloaded ones, music doesnt play etc). I finally gave up after 3 months and limited it to use LTE and have had no complaints since.

Huh, weird. I'm also on fi and I regularly get 500+mbit down/45-60 up, a handful of times I've gotten 900/80. Suburbs in major metropolitan area.

> I feel like signal is lower in general, and I run into a lot more cases where connections seem to hang entirely.

Exactly my use case.

I recently had to buy a new phone (broken mine) which has support for 5G. Thought that was worth a shot and enabled it on my SIM. I have now disabled it completely in Android and use 4G all the time, which works perfectly fine. I'm not downloading anything from the phone anyway and the most stress I put on its network is streaming music occasionally.

On tmobile, in the US, it’s definitely worse. I usually leave the cell settings to LTE on my iPhone. Occasionally, I’ll switch to 5G Auto. When I get the “uc” symbol next to the 5G, it’s often faster, but not always. I’d say I have a 30% hit rate on 5G where the speed is noticeably better. But as others have said, often I can’t get a web page to loaf even with the phone reporting full signal strength.

Try using Cloudflare warp. If it makes a significant difference then you know the problem is traffic shaping.

Don’t use fast.com on cellular as it’ll often get throttled.

Related to this, does 3G just not work anymore?

I’m in Ontario, Canada and whenever my phone loses LTE and shows full bars 3G, the internet doesn’t work at all.

Different continent (Europe) but something similar happens with my provider (Orange).

3G works perfectly to make phone calls (since VoLTE is a bit of a disaster here, most phones don't work with it) but for data is pretty much hit or miss. Some antennas seem to provide Internet and in others even ping (to 8.8.8.8 or 4.2.2.2) doesn't work.

For me I don't really need 5G at all as 4G already fast enough. I did not realize my phone was actually 5G until noticed the battery could barely make the day.

So I looked the battery statistics, and it seemed the radio ate lots of battery. It turned out that my mobile provider turned on 5G silently for me. So I forced the phone to connect to 4G or below and everything returned to where it was.

As the shift up in frequency happens to chase bandwidth, the penetration through structures and things like foliage, drops off a cliff.

This is what I've experienced. When the connection is working, the speed is great but it doesn't work at all inside the same buildings where I used to get a decent connection and speed before 5G.

In my experience "bars" of 5G don't have much correlation with actual (tethered) megabits. In some places it's like broadband speeds, and in some places I've had to explicitly switch back to LTE to get a decent connection even though my phone claimed it had a perfect 5G connection.

In my experience this applies to "bars" of all Gs, I've been in many places with full bars and unable to load anything. Seems that bars don't indicate backhaul bandwidth or congestion.

Haven't they always been signal strength? Which is only an issue when it is really low.

And with 5G we are dealing with so many frequency bands, users and so on that the strength of signal really tells little.

It's not even true signal strength. Certainly on 2G (I'd have to double check on later technologies) the radio can inform the UE to treat the recieved signal level as being an arbitrary amount better than actually measured.

This applies to signal level as reported to the user as well as internal metrics, eg handover (is this other cell better, should I switch to it?).

I’ve noticed this a lot with Verizon 5G UW (as my iPhone labels it). It will seem as if my entire connection is frozen (no pages load but they don’t error out either). If I turn airplane mode on and back off again 9 times out of 10 I immediately get a usable connection.

I don't have it on my phone but I recently got a home 5g modem and it's excellent - speed and latency are both good and indistinguishable from my old cable internet service. If nothing else, I welcome it as a long-overdue competitor to all the awful ISPs.

5G is so much better than LTE here in Australia, the difference in latency alone is really great.

Australian mobile data reception and speed is insane. Living in the UK really makes me appreciate how reliable and fast our 4g and 5g is. Part of that is not having impenetrable old stone buildings, but even outdoors the difference is huge.

We kinda need it tho to help partially make up for the absolutely abysmal copper NBN.

Yeah the 4(LTE)/5g is mostly pretty decent in well populated areas, I tend to get between 400-1100Mbit/s on my iPhone.

It took me 9 years living in Melbourne before I could get (at any cost) almost he same speed internet I had back in NZ 10 years earlier. Australian internet and politics are both very disappointing.

I turned 5G off on my phone and my data speed literally doubled. So I'm getting better battery life, better signal, better speed, and a bonus of less congestion (I assume, I'm not a networkologist) on the LTE network because of all the phones that default the poor suckers to 5G.

Anyway, I'm not all that thrilled with 5G. But I'm a bit of a luddite, I guess, because I also don't want to live in a world where EVERYTHING is connected. I don't want my car to have an internet connection. I don't want 5G boots. I don't need a 5G rain poncho.

Of course they want to push immature technology if it means billion phones will be set to be in need of upgrades.

We all know high bandwidth and low latency are only required on very specific needs.

We don't play e-sports and play 8k videos on phones.

My experience is the same as yours. If I leave 5G turned on my phone insists on staying connected to a 5G network even when that speed is less than two megabits per second in an area where 4G speeds exceed 50 megabits. It takes it the better part of a minute with no signal to fall back to LTE, which means constant connectivity loss when driving. Every six months or so I turn 5G back on to see if coverage has improved but it never has.

5G only really shines in super crowded situations like sports stadiums and music festivals.

In normal everyday situations 5G just drains more battery and causes more glitches due to the more frequent handoffs.

It's also much faster than 4G, of course, but it's quite rare for me to need speeds above 4G on my cell phone. I'm usually not downloading large files when I'm out and about away from WiFi.

You seem to be confusing the set of standards called 5G and the high-frequency mmWave bands that are an option part of it. mmWave makes little sense except in special areas and that's also why there's very very limited deployements of it. On 5G as a whole? Lots of network-level architecture and signalling improvements still, but depends on NSA/SA, used bands, implementation quality, end-user equipment bugginess, etc.

In my experience, there are areas where 5G coverage is much more spotty: the phone in one location has 5G and move it half a foot to the left or right and it’s got 3G. There are other areas where 5G is everywhere. 5Gs speed way faster than 4G in my experience. Although I think a lot of the speed you get is artificially imposed caps by the cell service providers.

I run into the ‘hung connection’ issue all the time with an iPhone 13 Pro on Verizon. It’ll just sit there and never load anything - to get network back I have to toggle airplane mode and force a reconnection. It does seem to be related to 5G - if I disable it entirely in the settings I don’t ever notice it happening.

I always thought of 5G being mostly for the carriers instead of being for the customers. The customers realistically don't gain all that much, but the carriers will be able to scale further. Even more customers will be able use the same service some already used. All the marketing fluff was what it was: marketing fluff.

I turn it off on every new iPhone purchase. Next time I’m in a sports stadium I’ll turn it back on … to check HN, naturally.

Glad I’m not the only one who thinks my cell speeds have gone down.

Maybe the tech needs time to mature? Did the switch happen too early?

5g is like 50 different frequency bands

Whereas 4g was like 1 or 2 frequency bands dictated by your carrier

I agree that the user experience is worse.

shorter length of coverage gives extremely high bandwidth, while a longer range of coverage gives 4g levels of bandwidth or maybe worse

hopefully the chipset gets better, I hope this isnt a fundamental limitation

Also battery life. Was in Florida a few weeks ago staying in a house without wifi but with weak 5G signal. My phone’s battery could barely make it through the day (which never happens to me otherwise). Turning off 5G (i.e., LTE only mode) seemed to help some.

5G download has been the same or worse than 4G, and signal can be much worse indoors/etc. When I get signal, 5G upload is usually better. 5G signal at home is better than 4G was, but I think that's because of more bands/towers, not because 5G is better.

idk. about US in some EU countries it's pretty common to only have LTE+5G sending stations.

I.e. most parts are hand over LTE and only some traffic gets switched over to 5G (e. g. video streams or large downloads).

the reason is simple that:

- backwards compatibility for all the non 5G devices is needed

- for most daily use LTE is enough

- use-cases which really need 5G and are in the normal phone radio network are still rare, the most common probably bring 5G to WiFi router in some cases. But such solutions are often seen as last choice if they is no other solution so mostly in the country side where the 5G/LTE net often isn't grate either (of available at all)

Latency and throughput in London appears to be largely the same.

I was hoping for something in the 20ms range for latency, but it always tests as in the 60ms range. Raw throughput seems largely the same as with LTE (ranging from 40-80mbps)

But if a underwhelming waste of time it seems.

I went from a NetGear M1 4G MR4100 router to a NetGear M2 5G MR6150 router and it is noticeably worse. The router overheats more easily, drops connections, needs to be rebooted.

I don't know if it is the product change or the 5G, but it is really frustrating.

I recently upgraded from an iPhone X to an iPhone 14 Pro on Verizon and did not see any improvement going from LTE to 5G. If anything, the connection became less stable and I decided to disable 5G and it works a lot better!

Looks like we've reached the feature saturation intersection, where the added benefit curve meets the curve of ridiculousness. This happened with razors too, who also coincidentally stopped at 5.

In a busy square in Portugal I got 1.1gbps and at a cafe in Malta I got 750mbps on 5g. Didn't check latency or packetloss etc, just a raw speed test in places I could see it was

My phone still doesn't support 5G. But a coworker is always complaining that when he gets 5G signal, comms work noticeably worse than when 5G signal is not available.

I've definitely seen that any time my phone says 5G UC that there is no connectivity. I have to reboot the phone, and it starts working again. Pixel 6.

Can't tell the difference frankly.

Then again UK doesn't have the sub ghz stuff which I imagine is responsible for most of the handovers/hangs

Depends on the provider and which bands they are using for it. At least on verizon in my area it’s strictly worse

It entirely depends on your carrier: frequencies used, number of antennas, backhaul, sharing w/ 4G, ...

The goal of 5G is to sell equipment at all levels because it has one more G. Mission accomplished!

Locally 3g was the most reliable. I think we're working backwards in accessibility of the network. It seemed just a few years ago, all we needed was a tower every so often, now we have towers and microcells. Is it a capacity issue or have we pushed into a spectrum that is more suited to dense cities rather than less dense towns?

I think capacity is a big part of it? People consuming a lot more video at higher resolutions so while 3G might have been fine for the time we needed to upgrade to squeeze more bandwidth into the available spectrum. It wouldn't be so bad if phones were better and switching to the best available signal, they they could use the efficient 5G connection when it's available, but fall back to 4G when the 5G is blocked by a building or something (which I gather is more common due to the high frequencies used)

I remember in the 3G era needing an Optus "HomeZone" at our house in Australia. Only 30 minutes drive from the dead center of a capital city, but our "country town" had a single cell tower and we were on the edge of town. So we'd get a single bar indoors and 2 bars outdoors

This isn't about your benefit ;)

5G should in the long run be cheaper for telcos to operate.

I have super fast broadband thanks to 5g. It's just little more expensive

Paranoid mode on

If 5G was forced into us so badly, chances are that:

- 4G was bad for the health - 5G allow much finger user tracking capabilities

Other than that, the rest of use cases were BS, as we already know.

More over in my country the telecoms had capped the 4G speed to half its speed and nobody complained.

5G drains the battery way too fast, and 4G is good enough for me.

Welcome to 2023. Just don't ask what's in store for 2033.

It's not just you. I have it disabled most of the time.

It's for sensing

5G is always terrible compared to LTE for me too.

"640K ought to be enough for anybody."

Yes.

Reply on news.ycombinator.com