Qualcomm's Simulated 5G Tests Shows How Fast Real-world Speeds Could Actually Be (theverge.com) 61
At Mobile World Congress, Qualcomm demonstrated the real-world potential of 5G by sharing findings of extensive network simulations it has conducted over the past several months. From a report: Instead of just offering guesses as to the gigabit-plus speeds that 5G technology could one day offer, Qualcomm's tests modeled real-world conditions in Frankfurt and San Fransisco, based on the location of existing cell sites and spectrum allocations in the two cities. The simulations factor in conditions like geography, different user demands on the network, a wide spectrum of devices with various levels of LTE and 5G connectivity for different speeds in order to accurately give an idea of what to expect when these networks launch. Additionally, the simulations are intended only to show the kind of 5G NR (New Radio) networks that could feasibly exist next year -- the non-standalone networks built in tandem with existing 4G LTE technology, not the truly standalone 5G networks that will come later on.
The Frankfurt simulation is the more basic network, based on 100 MHz of 3.5GHz spectrum with an underlying gigabit-LTE network on 5 LTE spectrum bands, but the results are still staggering. Browsing jumped from 56 Mbps for the median 4G user to more than 490 Mbps for the median 5G user, with roughly seven times faster response rates for browsing. Download speeds also improved dramatically, with over 90 percent of users seeing at least 100 Mbps download speeds on 5G, versus 8 Mbps on LTE.
The Frankfurt simulation is the more basic network, based on 100 MHz of 3.5GHz spectrum with an underlying gigabit-LTE network on 5 LTE spectrum bands, but the results are still staggering. Browsing jumped from 56 Mbps for the median 4G user to more than 490 Mbps for the median 5G user, with roughly seven times faster response rates for browsing. Download speeds also improved dramatically, with over 90 percent of users seeing at least 100 Mbps download speeds on 5G, versus 8 Mbps on LTE.
Wow (Score:5, Funny)
Frankfurt is in Germany (Score:1)
TFS doesn't only mention San Francisco, but also Frankfurt, which is in Germany, which it self is here in Europe - i.e.: a continent where not all ISP put as absurd limitations on data bandwidth as you have to put up with your USAmerican Telcos.
Re: (Score:1)
Seriously, how can you Europeans be so insufferably smug all the time? Need I point out Europe is the place that charges to use public toilets and for water at restaurants. They also created ezjet and ryanair. Nobody does cheap like Europe does.
Re: (Score:3)
Need I point out Europe is the place that charges to use public toilets and for water at restaurants.
That's because we famously hate socialism, of course.
Addendum: data points (Score:3)
addendum:
O2 [o2online.de] (an example of ISP I used during my stay in Germany) has currently offers of 25 or 20 GB per month.
At currently simulated 490Mbps (roughly 60MB/s) it would take between 300s (5 min) to 400ms to max it out, around 10x more that the above 30s example.
Also, once the limit is hit, the device isn't cut off internet, the speed is simply degraded to 1Mbps.
There are other countries in Europe where it's not even customary to have data limits : Switzerland is an example thereof (on most non-pre-paid-plans
Re: (Score:1)
This sounds like TMobile US with lower limits and harsher throttling.
TMobile gives 28gb, and then only deprioritizes during busy times (full speed most of the day).
Re: (Score:1)
Meanwhile, US custommers will probably have their monthly limits increased from 1GB to 2GB.
I've had a 16GB limit for a couple of years now, and I could choose a higher plan if I desired. As it stands, I don't need even what I have and should probably go to a lower plan. Anyhow, your assumptions regarding US limits are misinformed, probably you are taking too seriously what you read on Slashdot.
Re: (Score:1)
Re: Wow (Score:2)
Re: (Score:1)
For reference: No one with a brain uses that "iB" shit.
Ever since we started counting bits and bytes, we've counted 1024 of them as a Kb or KB. Similarly, we use 1024^2 for Mb and MB.
These are not SI units and never have been. The presence of the b or B means there's no ambiguity or confusion (unless you're a retard).
There are 2 main areas where people have gotten this fucking wrong despite the fact that they knew (or should have known) better.
1: Storage manufacturers. They knew what they were doing yet
Re: (Score:2)
So, your point is that 1kB is 1024 bytes and 1kbps is 1024bps except when talking about networking or storage. Got it.
I promise to use binary prefixes when I am not talking about networking or storage.
Re: (Score:2)
Ever since we started counting bits and bytes, we've counted 1024 of them as a Kb or KB. Similarly, we use 1024^2 for Mb and MB.
These are not SI units and never have been. The presence of the b or B means there's no ambiguity or confusion (unless you're a retard).
This war was lost many years ago.
1 MB is 1000000 bytes.
1 GB is 1000000000 bytes.
End of story. If you assert otherwise or attempt to ship product with other definitions your customers will NOT be impressed. They will think you're a retard.
Re: (Score:2)
Re: (Score:3)
I wouldn't have thought this would make a big difference, but it did show up as GSM phones having slightly better talk time than CDMA phones. GSM uses TDMA for voice - each phone is assigned a timeslice and can safely turn its radio off outside of that timeslice. In CDMA, all phones transmit and receive at the sam
Re: (Score:2)
Those speeds start to compete with land-based internet service, which create some interesting possibilities, like....
Mobile 5G could bring relatively fast internet service to many rural areas that are still starved of high speed internet. It would be much cheaper to build a tower to service many square miles than laying and maintaining cable/fiber down all rural roads.
"Low-bandwidth" home internet users (i.e. basic email and web-browsing) might forego the standard cable internet bill for a consolidated mobi
Re: Yay, more speed.. But for what? (Score:1)
More speed so that websitescan add more JavaScript bloatware/malware. Usersâ(TM) hardware will be more accessible for third parties and, in return, we will keep 1990s browsing speeds.
Re: (Score:2)
It also means that I can get to the Slashdot SSL error pages faster. Yay!
Re: (Score:3)
Yeah, the "real world" test is when the carriers dilute the number of available cells, stifle the backhaul, choose frequency bands that only a mother could love, then find ways to choose phones that aren't really capable of chewing through downloads quickly.
It's a nice proof of concept, but the real world is an unforgiving bear.
Re: (Score:3)
Re: (Score:3)
I can't quite dig up *that*, but there is some content around the same time period talking about '144 mbit' versus the 300 mbit theoretical max, so it seemed that about the same time in development, they were saying 'about half'.
Now this is saying the about half was too optimistic, and instead 1/6th turned out to be the case. If the same carries over to this, then we would still be talking about 150 mbps. In other words, what was hyped for LTE 'real-world' may be the '5G' real-world.
Of course, this *could
Re: (Score:2)
To get 144Mb you need multiple antennas and multiple channels. It's available in Japan but you need a special mobile router (phones have only one antenna) and a special extra expensive SIM card. It's marketed to business users and I doubt you get that full speed, but it should be faster than single antenna.
Re: (Score:2)
I'm in Canada, on fixed wireless with a hub with a couple of antennas, dual channel (2.4 and 5Ghz) and the best I get is 15Mbs down and now that all my neighbours (perhaps 100-200) seem to have the same setup, it is currently 5.5Mbs down and 1 up, last evening it was 1 down and 3 up. Not very impressive, though it saves the ISP money in just building a cell tower with some government funding instead of bringing the fiber all the way out.
FTFS... (Score:2)
"roughly seven times faster response rates for browsing"
What do you suppose this means? Latency dropping from 30-60 mS to 5-10? Page starts loading seven times faster? Do slow sites magically load seven times faster?
Not unambiguous.
Initial fonts, styles, and scripts (Score:2)
Perhaps the most honest metric is the time to first meaningful paint [google.com].
When the user navigates to an HTML document, a browser doesn't immediately display the data as it comes in. Doing that would cause the layout to jump around as style sheets, images, and fonts provided by the server replace those built into the browser and operating system. This jumping is often called the "flash of unstyled content" (FOUC). So before rendering anything, some browsers wait until the layout "above the fold" (that which can b
Still not even 4G (Score:5, Interesting)
Browsing jumped from 56 Mbps for the median 4G user to more than 490 Mbps for the median 5G user
So, we'll still be waiting for actual 4G speeds. Maybe "6G" or "7G" will finally meet the standards for 4G.
https://en.wikipedia.org/wiki/... [wikipedia.org]
Re: (Score:3)
Peak speeds are a pointless measurement anyway. What users care about is if they are in the town centre how congested is the network and how bad will their browsing/app experience be.
Latency is the biggest factor. If every packet gets delayed by hundreds of milliseconds due to congestion then it takes much longer to open web pages, get map data, pull down emails etc. The back-and-forth packet exchange is where what makes it slow.
Re:Still not even 4G (Score:5, Informative)
However, this assumes that data use is fixed and we know from history that as more bandwidth becomes available, consumption increases as well. Eventually though we're likely to reach a point where demand for more data doesn't scale in step with availability, but I think that this will help out considerably until people find new ways to consume mobile data.
Re: (Score:2)
Eventually though we're likely to reach a point where demand for more data doesn't scale in step with availability, but I think that this will help out considerably until people find new ways to consume mobile data.
All I know is that my communication demands as measured in "fractions of cell taken up by serving me on average" has gone down over the years. It might be because I am getting older, of course.
I used to be able to keep a significant fraction of a cell busy, with either multi-channel circuit-switched data or GPRS over plain GSM. Even with EDGE, I could put measurable load on a cell even with just one phone. Today on 4G my meagre usage drowns in everyone else's.
Re: (Score:2)
I think we've already reached that point. I can't think of anything a typical phone user would want to do that's more bandwidth-intensive than streamed
Re: (Score:2)
Don't need faster. Need more reliable. (Score:4, Interesting)
Re: (Score:2)
Sub 6GHz band (Score:2)
Browsing faster than download speed? (Score:2)
My understanding of "browsing" says that it is a series of file parsings and downloads, in a sequence, interspearsed with various short or long pauses. That suggests that the average browsing speed should be less than the peak file transfer speed.
In the TFA, the "median 5G user" experiences a 490 Mbps "browsing speed", while the "90%" experienced a "download speed" of 100 Mbps.
How could that be? Is "downloading" capped to the nice, round number 100?
Re: (Score:2)
I assume the test involves some simulation of the congestion control policies that Qualcomm expects cellular ISPs to apply. An ISP can oversell capacity more for bursty interactive use than for bulk downloads.
Re: (Score:2)
But doesn't that kind of defeat the purpose of demonstrating the technology? The figures would demonstrate the ISP limits, not the technology limits.
I thought Qualcomm would be interested in demonstrating the technology as such.
For a given subscriber-to-tower ratio (Score:2)
The summary implies that the test assumed current subscriber-to-tower ratios. I imagine that prioritizing bursty interactive traffic improves overall user experience for a given subscriber-to-tower ratio.
Who cares? (Score:1)
Maybe I'm extraordinarily lucky. I don't know, but I don't struggle for phone service, basically ever. Verizon 4G LTE is "good enough". I watch Youtube videos casually. Navigation works. Texting, music, installing apps wherever, etc. There are problems but they are rare enough that I can't actually remember the last time I had a service problem.
The only exception is when hiking backwoods trails and half the reason I do this is to get OFF the grid. Service in Italy was basically useless, calls only. But serv
Re: (Score:3)
I personally want cheaper data rather than faster data. I actively avoid data because my phone is basically direct tap to my bank account for these vampires. I want a fair amount of data for a fair price at a decent speed. Speed today is more than fine, but I hate feeling like I am getting robbed every time I do much of anything on my phone.
Not a demo (Score:2)
Re: (Score:2)
Why Bother (Score:2)