I mean don't get me wrong, I continue to find it remarkable that denser constellations such as 256QAM and 1024QAM work at all, let alone still having usable Wi-Fi throughput at -70dBm (0.000,000,000,1 watts).

In reply to a post by jchamier:

Only possible with hardware that has 2.5gb ports or aggregated 1gb ports to the wired network and using WiFi-6 but I read that you could achieve 1.3 Gbit over radio. I don’t have a reference. Perhaps I’ve completely misunderstood ?

Well, Cat6 cable will give you 10Gbps up to about 55 metres and Cat6A up to 100 metres - provided that the ethernet ports support that rate. You can see real world tests of Wi-Fi 6 systems, including mesh here. Speeds of 1.52Gbps appear to be possible if you're within five feet of the access point but if you put a bit more distance or obstacles in the path you'll see less than 1Gbps.

Higher speeds on Wi-Fi 6 require 80MHz or 160MHz channels and of these only one 80MHz channel doesn't fall into the DFS category in the UK [Source (PDF)]. Whether you can use DFS channels or not depends very much on where you're located. I use one at home without any issue but the only time I tried to use 5GHz for a mesh wi-fi link in deepest rural Worcestershire it turned out to be a disaster with constant connection drops. I then discovered the installation was less than ten miles direct line of sight from a Met Office radar installation co-located with a NATS air traffic radar at Tittersone Clee Hill.

There are no circumstances at the moment in which I'd consider using a wi-fi link between access points when I could be using ethernet.

In reply to a post by prlzx:

I mean don't get me wrong, I continue to find it remarkable that denser constellations such as 256QAM and 1024QAM work at all, let alone still having usable Wi-Fi throughput at -70dBm (0.000,000,000,1 watts).

Agreed, these are impressive technologies. WiFi seems to be benefiting from progress in the mobile industry.

In reply to a post by caffn8me:

Well, Cat6 cable will give you 10Gbps up to about 55 metres and Cat6A up to 100 metres - provided that the ethernet ports support that rate. You can see real world tests of Wi-Fi 6 systems, including mesh here. Speeds of 1.52Gbps appear to be possible if you're within five feet of the access point but if you put a bit more distance or obstacles in the path you'll see less than 1Gbps.

Thank you! That sort of test is what I was looking for, I was going through the Small net builder site's test history, but couldn't find anything useful.

Whether you can use DFS channels or not depends very much on where you're located. I use one at home without any issue but the only time I tried to use 5GHz for a mesh wi-fi link in deepest rural Worcestershire it turned out to be a disaster with constant connection drops. I then discovered the installation was less than ten miles direct line of sight from a Met Office radar installation co-located with a NATS air traffic radar

That must be very frustrating. I'm using 80 MHz DFS channel here without problem, and I live close to an airport, so I guess it depends on the configuration of the radar.

There are no circumstances at the moment in which I'd consider using a wi-fi link between access points when I could be using Ethernet.

I understand, but for some people there is no easy way TO install a cable, but it seems we are NOT close on performance, yet.

At the moment the only real way to know how well a particular system will work is to buy (or borrow) the kit and try it.
Some decent educated guesses can be made based on specifications and materials.

As per the document shared by caffn8me Ofcom were consulting on reducing the DFS requirements (particularly indoor) and widening the available channels to include 6GHz but to avoid having manufacturers having to release country-specific equipment (such as in USA) UK tends to try to co-ordinate with the rest of Europe.

This will likely remain true (even now) as people will still want to be able to use the same phone on Wi-Fi when they travel.

The exact regulatory rules in effect are applied by the firmware in concert with a user-configurable country-code in the AP as that determines what channel(s) are available.

But manufacturing still needs to design and build radios, antennas and chipset that operate just as well in the new frequencies so agreeing a common min to max operating freq range over wide geographic areas still helps in terms of what models will be sold in what economic areas (similar to what happens with GSM / CDMA versions of the same smartphone now).

And at 80MHz having 12 distinct usable channels instead of a single non-DFS channel would make a difference once it filters through to end-user devices.

Edited by prlzx (Tue 05-Jan-21 21:46:28)

Thanks, and yes, agreeing with the rest of europe standarised frequency ranges seems eminently sensible (and I believe that relationship started before the 1970s with TV transmissions?). Completely agree rest of your post.

Technology to watch!

In reply to a post by prlzx:

I mean don't get me wrong, I continue to find it remarkable that denser constellations such as 256QAM and 1024QAM work at all, let alone still having usable Wi-Fi throughput at -70dBm (0.000,000,000,1 watts).

For the benefit of any readers who, like me, find it easy to lose track of zeroes that's 100 pico-watts