Indian telecoms leaps from 2G, to 4G, to 6G – on a single day India yesterday laid out its ambitions to become a big 6G player – on the same day its biggest carrier tried to lift millions from 2G to 4G. The 6G push was announced by the Department of Telecommunications, as it launched the Bharat 6G Alliance that aims to "support and energize Indian participation in standards development …
The greatest example yet of a solution looking for a problem.
What the heck problem is 6G going to solve that mmwave 5G does not? The standards people are pushing it because they have to push another generation or they'll be out of a job. The incumbents with big cellular patent portfolios are pushing it because those LTE patents don't last forever, and they need to create an illusion of inevitability about 6G so that by 2030 consumers will believe if they buy a phone that lacks 6G they might as well be using a landline.
There may be some niche commercial/industrial uses for 6G (though I struggle to think of any) but no one reading this will ever have a use for it in their lifetime. Any improvements they make for special features that are irrelevant to cell phones but might be useful in certain commercial/industrial settings (e.g. connection times measured in us instead of ms) could be added as extensions to 5G, similar to what LTE Advanced was to LTE. But no, they want to give us a full boat new standard, including using spectrum into the 100s of GHz so they can claim tens of gigabytes of throughput which no personal device could possibly need or use.
I look forward to the inevitable "640K is enough for anyone" replies from bozos who can't think of any reason for 6G to exist either, but just assume because we needed more than 640K we'll need more than 5G - ignoring that apocryphal 640K story was from very early in the lifespan of PCs, when cellular was 0G (i.e. pre standards analog) 640K may not have been enough for everyone, but the 64 bit address space we all have now in our PCs and phones sure is!
" the inevitable "640K is enough for anyone" replies "
At the beginning of tech lifecycle, things are changing very quickly. 640K was clearly not enough for anyone, and as tech improved it very quickly went to MB, hundreds of MB, GB.... Now I wouldn't say 8GB or 16GB are enough for anyone, but 8GB of AM probably covers 50% of users worldwide and 16GB probably covers 90%+ of people whose use cases are browser, email etc (also glossing over the fact that the servers serving them content will be using far more resources than just the local user PC). But over the next decade, the growth will be basically flat, driven mostly by increased OS "requirements" that are not really required*.
Similar with phones, the difference between every successive generation of wireless technology and the phone technology itself gets less and less every year, and will only jump with a major breakthrough that suddenly ups the requirements.
*The only scenario I see in the near future where PCs / laptops will suddenly require much more horsepower is "AI" (large statistical models) that can be locally driven (which is preferable due to privacy reasons). ie I would never have an Alexa or equivalent device listening in on me 24/7 and sending the data to Amazon or whoever for processing. I would, however, be very interested in am 'intelligent'** personal assistant that can process my commands locally and act locally (intelligent home automation) and only do online tasks when required and specifically ordered.
**Not really 'intelligent', but able to recognise my voice and do a few simple tasks
Now that we have devices that have multiple Gb in RAM, it seems to be the case that 'Device has Xgb so we (user/developer) may as well use as much as we can rather than craft something non memory hungry". This now seems to be the case with mobile. From what I've briefly read both 5G and 6G will need more base stations than the prior generation, so really only cost effective in urban areas, which doesn't square with one of the perceived use cases of 6G being autonomous vehicles. I can only think that a meaningful use case is to supplant fixed (fibre) broadband to premises (FTTP).
> may as well use as much as we can rather than craft something non memory hungry
That, and adding more advertising and tracking bloat on every page, since connection speeds are faster today than 10 years ago. Oh, and a side dash of auto playing video or two, for good measure
This whole 5G/6G requirement is driven by the exponential increase in RAM which then leads to developers "may as well use as much as we can rather than craft something non memory hungry" attitude which leads to bloated apps that need more memory and more bandwidth to download them - hence the need for 5G and 6G going forward. Do you know that the Facebook app on the App store is now 400MB in size, so you need a very fast 5G connection to download it in under a minute? The same app was under 25MB 10 years ago, and not much new has been added since.
When I look at my day to day usage and how its changed over the years I'd probably need a bit more than 640KBytes --- but not a whole lot more. The reason's straightforward. Most of my system's resources are consumed by the web browser and much of that is providing 'analytics' (spyware) and delivery of unwanted advertisements. Web pages are also greedy in both system resources and network bandwidth -- a few years ago an Australian developer delivered an excellent presentation built around the idea that a typical web page is longer than a giant work of literature such as Tolstoy's "War and Peace".
We tolerate this situation because we have to. With mobile the only reason for lightning speed and ultra low latency would be the need to cram more and more users onto a network. This seems reasonable except when you realize that 'network' to most developers involves web protocols which are about the most inefficient way of transferring information between two points that anyone has been able to devise. (Very few developers have a clue what happens when they shell off 'curl' to get a file -- sure, its convenient but its the wrong kind of convenience for both the user and the network -- it should be "for occasional use only".)
"I can only think that a meaningful use case is to supplant fixed (fibre) broadband to premises (FTTP)."
You'll never supplant fibre with a shared medium resource.
There are a small handful of cases where premises are sufficiently distant that it might be a reasonable option, but that runs completely counter to the probable coverage of the base stations.
In those cases Starlink already provides a decent (but still well behind fibre) connection...
Smaller cells can already be achieved by simply lowering the power of the transmitters, so you could pack an urban area with more cells if you were so inclined.
Autonomous vehicles is an appalling use case - you want vehicles to be as self contained as possible, though a rapidly negotiated local network could be useful - I'm approaching this junction, and can talk to the few dozen cars within 400 yards to share plans.
> "if they just got 4G working consistently..."
I, Personally, would _love_ to see the carriers (looking at _you_ Comcast)
provide voice via 4G LTE with fewer (preferably 0) cases of my own speech
from 1-10 or so seconds ago competing with the other person on a call.
Halfway decent echo suppression has been around for at least 50 years,
with far less hardware.
Yeah, I know, it's no longer a hardware problem, as pretty much everything
has been punted to software. We all know software is perfect, right?
It supports mmwave and also greatly improved spectrum sharing in situations with a lot of contention. I think there was enough new to deserve +1 'G', but they've reached the limits of what will benefit consumers with 5G. Anything left are niche cases that don't justify Apple and Samsung bothering to support 6G - and paying Qualcomm et al patent license fees for something their customers don't need.
whether the phone service shops at airport entry points will start offering complete phones to temporary travellers instead of just SIM cards? Might be handy; I've never really wanted to mess around swapping SIM cards on arrival and tend not to use my phone at all (except on wifi) rather than be charged the frankly extortionate overseas rates both my UK and DE suppliers thing I'll be delighted with.
Not necessarily, you could just rent the phone on arrival and deposit it into a bin on departure.
The real problem is, if you're using a temporary phone, how do you get any contacts etc onto it from your phone, and how easily can you erase any temporary documents, history etc before giving it back?
When travelling to a country where roaming isn't included in my plan I take an older phone with me to put the local SIM in and then use it as a hotspot for data,
If you have a phone that supports an eSIM you can now get an app that will sell you a local 1 week, 1 month etc plan via the eSIM (not used it yet as I haven't been outside the EU since I have had a phone that supports eSIMS)
Got an eSim for iphone when I went to US on holiday thinking it would also give me a temporary US phone number. It didn’t. :-(
Probably just chose wrong network but none of the other companies i looked at later offered them either.
For normal data use in US a 3 payg sim with some data isn’t a bad option but obviously issue with U.s phone number still applies.
I'm trying to think of the notion of "temporary travellers" in somewhere like India. The overwhelming majority of travellers will be Indians and there will be shops selling phones, just go in one of those. Of course, even if you get a phone that reduces your costs, you'll be increasing the costs for anyone who wants to call you.
Wouldn't it be great if mobile phone companies didn't price-gougle consumers with their stupid roaming charges. It would be wonderful if you could just use your phone when you landed in another country and not worry about the cost. All this talk of renting a local phone, or buying a local SIM card is the wrong answer. This doesn't require a technical or technological solution.
I can understand some kind of surcharge but it doesn't need to be anything like as much as current rates. Think of some popular holiday destinations that get swamped by foreign visitors: some kind of charge would help them invest in their networks. The problem with roaming charges is that the "home" networks like to levy a charge and effectively auction their subscribers to the roaming networks.
Mind you, here in Europe we've largely forgotten about roaming charges thanks to those dastardly socialists in Brussels! :-D
I'm not aware of any UK network that allows you even to use WiFi calling when you're overseas and, indeed. most go out of their way to frustrate simple-minded workarounds like VPNs. Unlike roaming, this would be entirely cost neutral as far as the carrier was concerned so there's no rational excuse for it except milking the customer.
There are networks in other countries that actively encourage WiFi calling abroad, so it's not a universal fetish.
"It would be wonderful if you could just use your phone when you landed in another country"
Absolutely agree but that still wouldn't solve the need for a local phone number when travelling. You'd be supprised how many people in Australia, the US or Africa are reluctant to pay for a call to the UK just to tell me my taxi has arrived (many can't even call internationally outside their own country).
A local physical SIM is pretty much an essential and always the cheapest option when travelling, something that ESIMs can't solve as (apart from being as rare as hens teeth) the standard requires you already have a working data connection (ie have already paid someone else for a physical SIM or roaming data) before you can activate one (scanning the QR code just downloads something from the internet - the QR code is not the entire SIM).
3G was actually enough. Turning it off just created a load of e-waste. Most people only upgrade their phones when their 'G' is turned off. If the industry gave a toss about the environment (which they do not) they would run with removable batteries and TRX modules, so the same phones could just shift up a G.
Just use R for Rupee. And Kiev, Marathon, Turkey, Bangalore etc. Would you prefer to call Japan 'Nihon-koku' and only ever write it as 日本国?
"3G was actually enough"
3G is generally regarded as a failed standard. Circuit switched voice calls with the assumption these would include low bitrate video, instead of being packet based wastes huge amounts of capacity and results in voice call quality issues due to CODEC limitations. Then you have the issues around "cell breathing" and the need for "transcoder free operation" to impliment HDvoice.
2G is outliving 3G in many places, as 3G just never really worked properly. Even 3.5G (HSPA etc) can't solved those fundimental issues.
5G and 6G are not about user level technologies. They’re about standards and IP ownership. Neither India nor China will sign up for telecom standards whose IP is foreign owned.
They’re going to craft their own because each of them has over a billion cell subscribers and well over half a billion smartphone users, and can’t be paying others royalty for that much of a market.
India took time with ensuring 5GI was ratified, it wants to avoid such delays with its own country specific 6G. It has nothing to do with how pertinent 6G is as a user facing platform, it’s all about control over IP and the hundreds of billions in tech equipment around it,
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2023
