What's new

Does OFDMA Really Work? Part 2

  • SNBForums Code of Conduct

    SNBForums is a community for everyone, no matter what their level of experience.

    Please be tolerant and patient of others, especially newcomers. We are all here to share and learn!

    The rules are simple: Be patient, be nice, be helpful or be gone!

11ax testing is many levels more complex that what was done with earlier revisions...

HE frames on the UL are triggered by the AP, and this is done when conditions are suitable ( delta time < 400ns) - if not, the AP shouldn't even try if there is more than 400ns arrival time from the client STA's - frequency stability is also significant in that the delta center freq must be less than 350 Hz (yes, and that ain't much on a 5GHz channel)

Client STA power control is also a thing - as the AP needs to have all the OFDMA RU's arrive at a similar power level, and they have to all arrive in the assigned RU groups, otherwise the RU's, and possibly the entire frame, is lost...

MU-MIMO is also about arrival time, and the I/Q offsets - the AP has no idea of north/south, but it does know about the channel timing, and the sounding reports from the client STA's - and with 11ax - MU links are reciprocal - if conditions are not conducive for DL-MU, it's not likely that UL-MU is going to be any better...

Back in the 4G standards development era - LTE looked at this, as did 802.16m (Mobile Wimax) - LTE decided that the complexity of OFDMA would not have much benefit, and went with a SC-FMDA (also known as LP-OFDMA), which is basically OFMDA with a single RU group...

OFDMA in 802.11ax is an evolution of work done in 802.16m - and we all know how successful that was...

I wouldn't worry too much about OFDMA to be honest... there are other comms tech where there is benefit (DOCSIS 3.1 is a great example), but on WLAN/WWAN, it has not shown results that were promised...

OFDMA-RU.png


OFDMA-UL-spectrum.png
 
I wouldn't worry too much about OFDMA to be honest.
And that is my conclusion. The average user, even once they have a few AX devices is unlikely to notice a difference.
This new whitepaper Qualcomm sent me basically says you need a LOT of BUSY AX devices to see OFDMA benefits. Is even the Home Scenario (section 5.1) realistic?
 
Is even the Home Scenario (section 5.1) realistic?

Assuming perhaps that every client STA is WiFi-6 in the home - which is unlikely... 2.4GHz clients on a WiFi-6 AP will generally run in 2.4 11n (20MHz), and my mix of 5GHz is N/AC, with a couple of AX (as mentioned).

QC_WP_51.PNG

Because of this - there's little to no gain because of OFDMA as most clients are WiFi 4/5..

The busy office (5.2) assumes the same - all client STA's are WiFi 6, and the classroom scenario is even worse - how many schools have the budget to replace all their chromebooks with shiny new ones, even if there were affordable devices with WiFi-6

Don't get me started on WiFi-6E - I'm trying to be family friendly, and I have some pretty strong opinions there (and salty words ;) )
 
Last edited:
Well, see if you can restrain your self from the "salty" words. :)I'd like your take on 6E, please.

I'll be nice ;)

1) Dedicated Radio or the barrens of ISM band - one either has to implement a dedicated radio to access the 6E expansion, or kick all the existing WiFi-6/5 clients over to 2.4GHz

2) Automated Frequency Coordination - 6E is not the primary user in that band, so there's the "database" that the regulatory agency (here in the US, it's the FCC and whoever gets the contract to establish and maintain it) - that does smell a bit - more so than DFS did in the UNII-2 bands - it's supposed to be anonymous, but the data is there, waiting to be monetized and even hacked...

I can appreciate the expansion in the upper bands - would have been nice at some point in the past for FCC to open up ISM a little bit more, allowing 12, 13, 14, but that horse has left the barn...
 
1) Dedicated Radio or the barrens of ISM band - one either has to implement a dedicated radio to access the 6E expansion, or kick all the existing WiFi-6/5 clients over to 2.4GHz

And all is not entirely lost in the 2.4GHz band - which is probably worth another thread/post - I have ideas and an open call to the vendors that can help make that band more useful for WiFi 4, 5, and 6...

@thiggins - ping me in PM - this might actually be a useful article starter on the main site.
 
I read that when it first came out. Look at the # of devices in each scenario.
Fully understood - and in my "quiet" home environment - continue to have OFDMA disabled ... along with MIMO on my AX86U :).
 
I'd like your take on 6E, please

@thiggins - A year and a half later...

WiFi6e - if you can get it, and have enough clients to justify upgrading the Access Point, then do it.

By all means, do it - with the understanding that there are other things - like security - WPA3 is the only supported option there, it's either open or WPA3 - which has impact to mesh and other multi-AP environments that need to support legacy clients in the 2.4/5 GHz bands.

WiFi6e is pure - there is no legacy support for non-AX air interfaces, so the clients and AP's basically are a rebuild from the old 11 b/g/n/ac, and this is not a bad thing.

Interesting to note - Apple recently released their iPad Pro's with 6e support - and they've provided a switch to disable it. I have to wonder if this is due to their testing, and potential interoperability issues?


sfx
 
I have 2 networks.

OFDMA + MU-MIMO helped to improve latency in my "work" network. It's not an office, it's a group of Airbnb apartments sharing symmetrical 1Gb WAN.

On the ax only channel, there are over 100 active ax STA's, approximately 40 of those are real people streaming and gaming. The benefit was considerable, significantly lower pings under load.

On my home network (25 STA's), OFDMA had a negative effect on speed and latency.

Both networks use Asus AX92u

The bottom line for me is that the technology is overkill in a home WiFi setting. Gamers in your house will want the lowest latency possible, so leave OFDMA off.

These technologies are enterprise solutions to enterprise problems, marketed at home users that don't have the problem in the first place.
 
Last edited:
I have 2 networks.

OFDMA + MU-MIMO helped to improve latency in my "work" network. It's not an office, it's a group of Airbnb apartments sharing symmetrical 1Gb WAN.

On the ax only channel, there are over 100 active ax STA's, approximately 40 of those are real people streaming and gaming. The benefit was considerable, significantly lower pings under load.

On my home network (25 STA's), OFDMA had a negative effect on speed and latency.

Both networks use Asus AX92u

The bottom line for me is that the technology is overkill in a home WiFi setting. Gamers in your house will want the lowest latency possible, so leave OFDMA off.

These technologies are enterprise solutions to enterprise problems, marketed at home users that don't have the problem in the first place.

Is that for both OFDMA/802.11ax MU-MIMO with dl/ul setting applied and 802.11ac Multi-User MIMO activated?
 
I live in a big city with many apartments and 5Ghz band is congested.
Will wifi 6 help in my case?
Do OFDMA and other Wifi6 technologies help only on the same LAN or do they help to fight congestion from neighbouring networks?
 
Turned it on on my ax88u pro running 388.3 we have 6 wifi 6 devices my and her MacBook pro and two ipads and two phones haven't noticed difference but in your testing ax88u showed improvements so I turned it on ul/dl ofdma+mu mimo i also got an old apple tv ,4k first gen i run in the bedroom on 5ghz and i get 480mbps on speedtest so it's not adversely affected by the ofdma
 

Support SNBForums w/ Amazon

If you'd like to support SNBForums, just use this link and buy anything on Amazon. Thanks!

Sign Up For SNBForums Daily Digest

Get an update of what's new every day delivered to your mailbox. Sign up here!
Top