Graph of real world 4G performance/speeds

bealtine

Hucfqb2.png

oscarBravo

The X-axis is concurrent users?

bealtine

oscarBravo wrote: »

The X-axis is concurrent users?

yes

bealtine

Here's the original source, well where I picked it up from :


http://nbnmyths.wordpress.com/

Refers to Australia but applies equally here

watty

It's a generous graph, as you move it SLOWER according to key for signal quality.

So for the number of users evenly spread over a full cell area, the median user speed is about 1/4 or less the suggested speed.

It's also full duplex 20MHz + 20MHz. If you have 10MHz, it's 1/2, if you have 5MHz, it's 1/4.

Very old news: 2009
http://www.radioway.info/comparewireless/img10.html

Here many users see under 0.5Mbps
http://www.radioway.info/comparewireless/img11.html

watty

bealtine wrote: »

Here's the original source, well where I picked it up from :


http://nbnmyths.wordpress.com/

Refers to Australia but applies equally here

No, I think here is much worse

I thought the headline speed is 1/2 to 1/4 here, only 10MHz and 5MHz spectrum per operator.

You really need 3 channels of 20MHz Duplex to get a graph as good as that on a real network, 120+ 120 MHz. Who has as much as that here?

bealtine

These are the conclusions of the mobile industry, granted it's from 2009 just after the LTE standards were frozen . It is however FROM the mobile industry:

http://www.atis.org/lte/documents/Does%20LTE%20Do%20What%20It%20Says%20on%20the%20Box.pdf

Explains the effects of the shared nature of radio networks, in this case mobile.

(old hat for the user of this group but worth re-stating now we have all the "sooperfast nonsense in the media)

Includes key points like this:

"At any given instant, the cell’s spectral resource is shared between all active users" also has an excellent graphic on page 16 which shows why LTE will never delivers 30Mb/s or indeed any guaranteed speed...

AlanS181824

Very interesting. Thanks OP

justforgroups

bealtine wrote: »

yes

You mean the scale is individual users? Or is the scale per hundred, thousand, million...?

bealtine

justforgroups wrote: »

You mean the scale is individual users? Or is the scale per hundred, thousand, million...?

individual users