Qualcomm's Simulated 5G Tests Shows How Fast Real-world Speeds Could Actually Be (theverge.com)

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.

Comments

Wow

[ico2]

How useful, to be able to reach my data limit in just over 30 seconds!

Re:Wow

[Solandri]

The benefit is not just greater overall bandwidth. Faster throughput means the 5G radio and modem on your phone does not have to be turned on for as long, resulting in increased battery life.

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 same time, The coding for each phone is orthogonal, so you can separate its signal from all the others (like writing vertically and horizontally on the same paper - the letters are orthogonal enough you can distinguish which ones are vertical and which are horizontal, even though they overlap). CDMA turns out to be better for bandwidth allocation since unused bandwidth reduces the noise floor thereby increasing the SNR and bandwidth available to phones which are transmitting (in TDMA that unused bandwidth is wasted). But TDMA turns out to use less power because you know there are certain periods of time when you can turn the radio off.

Re:If it is simulated itâ(TM)s not real world

[postbigbang]

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:If it is simulated itâ(TM)s not real world

[bugs2squash]

I wonder what the simulated throughput of LTE was back in the day

Re:If it is simulated itâ(TM)s not real world

[Junta]

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* be a more realistic simulation and maybe there is a lot of headroom, but historically speaking that's never been the case.

Still not even 4G

[nasch]

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/...

Re:Still not even 4G

[AmiMoJo]

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

[alvinrod]

But the congestion is also tied to the throughput. If a user can transmit more data in a given time slice they can finish sooner (or require fewer time slices to get their data) which makes it faster for the next person to get what they want and they're able to get that a little bit faster as well. The faster you can clear out someone waiting, the more it keeps people from piling up.

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.

Don't need faster. Need more reliable.

[DogDude]

I don't need faster cell or wired Net service. I need more reliable. Faster without reliability is useless to me.

Addendum: data points

[DrYak]

addendum:

O2 (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, only speed is limited (together with minimal guaranteed speed), not total download volume)

I'm too lazy to do a systematic check but lots of European countries are likely to be in similar situation.

And that's today's number. By the time 5G finishes getting deployed to customers, the various plans will be adapted to it (probably with data limits in the 100GB range and higher speed limits / minimal guarantee).

Meanwhile, US custommers will probably have their monthly limits increased from 1GB to 2GB.

Re:Frankfurt is in Germany

[K. S. Kyosuke]

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.

Re:Who cares?

[Moof123]

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.