Is Anyone Using MU-MIMO?

[thiggins]

If you are using MU-MIMO with more than one MU-MIMO device, please share your experience.
Has it helped? Has it hurt? Does it make any noticeable difference?

 

[pege63]

I have MU-MIMO devices (Intel Dual Band Wireless-AC 8265 (2x2:2)) but i have not activated MU-MIMO on my router, it makes no difference wath i experiencing it.

 

[sm00thpapa]

My Galaxy S7 supports MU-MIMO and maybe a few more cell phones. I can't really say I see huge improvements while using my cell to watch Netflix or Youtube videos.

 

[thiggins]

My Galaxy S7 supports MU-MIMO and maybe a few more cell phones. I can't really say I see huge improvements while using my cell to watch Netflix or Youtube videos.

I would not expect a difference using only one MU-MIMO device. Only when using two or more simultaneously.

 

[sfx2000]

I would not expect a difference using only one MU-MIMO device. Only when using two or more simultaneously.

Where things can get odd - if the AP has MU enabled, some devices can get into odd behavior, since this is a IE they may not be expecting - some of the smartphones, we've had reports on the forums where they go from 2-stream operation down to single stream if MU == true

Similar to the 160/80+80 issue/workaround I recently posted...

 

[thiggins]

some of the smartphones, we've had reports on the forums where they go from 2-stream operation down to single stream if MU == true

That means it's a BRCM MU-MIMO chipset.

 

[sm00thpapa]

At this point I have no faith in Broadcom concerning MU-MIMO.

 

[ulaganath]

I just tested a mu mimo ac adapter with R78k. Difference was there but not actually practical. As its not two mu mimo but two ac adapter connected at same speed transferring locally. Mu mimo client dipped download from 28 MBPS TO 18 when non mu mimo client doing 100Mbps traffic on speed test

Sent from my ASUS_Z00AD using Tapatalk

 

[thiggins]

Folks. I appreciate your taking the time to respond. But unless you are using at least TWO MU-MIMO adapters at the same time, your feedback isn't helpful.

 

[pege63]

Is all MU-MIMO Router systen 4x4 11ac 80MHz BUT only 2x2 160/80+80MHz in MU-MIMO mode?
I have 3 MU-MIMO clients two laptops and 1 phone but when i use MU-MIMO mode the NONE MU-MIMO clients gets slower speed and star to lag, its that normal?
So the MU-MIMO is off in my router

 

[thiggins]

Is all MU-MIMO Router systen 4x4 11ac 80MHz BUT only 2x2 160/80+80MHz in MU-MIMO mode?
I have 3 MU-MIMO clients two laptops and 1 phone but when i use MU-MIMO mode the NONE MU-MIMO clients gets slower speed and star to lag, its that normal?
So the MU-MIMO is off in my router

What router are you using?

Devices that do not support MU-MIMO (SU) should be, best case, unaffected. The theory is that MU-MIMO devices should free up airtime for SU devices. So, if anything, SU devices should get more bandwidth.

But real life for MU can be very different.

 

[pege63]

What router are you using?

Devices that do not support MU-MIMO (SU) should be, best case, unaffected. The theory is that MU-MIMO devices should free up airtime for SU devices. So, if anything, SU devices should get more bandwidth.

But real life for MU can be very different.

Its a netgear R7800

 

[ulaganath]

the situation is very complicated for the two or more MU-MIMO to get advantage. As both client need to be in different direction say 180 degree distance. As it cant focus two clients at same location.

I have 2 mu-mimo. Yet to test it . WIll keep you posted.

One way to test is do a online and local transfer. As for online mine is limited to 100mbps.

So this might be good way to test. say

client 1 tv a 4k UHD playback from local NAS wirelessly with bit rate of 20+MB/s

Client 2 laptop with MU-MIMO usb 2.0 adaptor with max signal doing 25+MB/s

so close to 360Mbps bandwidth is consumed wirelessly.

Now if you do traffic of say 100+ mbps on wirelessly via SU- clients will tell if the first two client bandwidth is chocked or not.

 

[thiggins]

I have tested MU-MIMO devices at close range. Side by side or separated by 180 degrees makes little difference.

Simplest test is using two of same MU-MIMO devices.

Turn MU-MIMO off in the router. Run a speedtest on each device separately.
Then run both at the same time. Record throughput in each case.

Enable MU-MIMO in router. Run the same tests.

Compare the TOTAL DOWNLOAD ONLY throughput of the two devices in the simultaneous download test with MU off with it ON. MU on should yield higher throughput.

 

[sfx2000]

I have tested MU-MIMO devices at close range. Side by side or separated by 180 degrees makes little difference.

MU tends to be more interesting in time delays from the AP - remember - time is the fire in which all burn - and with MU - it's more the scheduling of delivery.

Beamforming and MU are more about time - as the AP really cannot tell one direction from another, but it can determine time delays based on reports from the clients. At the same range with the clients, it's really hard to do MU, because all the numbers end up being the same.

And just like beamforming in SU - it's really delaying the subcarriers in the baseband in a phase shift - which does tend to move the pattern around in the analog RF domain at the antennas.

 

[Razor512]

In thinking more about this topic, I wonder, is it possible for router makers to detect which clients support MU-MIMO and provide that data in the devices list?

Since many devices do not make it clear if they support those technologies or not (router makers are quick to list support, but you often will not see a smartphone maker list it on their specs page). I feel it would be good for wireless routers across the board to provide details on what is supported by the client devices within the web UI.

 

[avtella]

The Nexus 5X does and HTC M8 is advertised as supporting it as well and quite a few more Android devices. On the laptop side you have the Intel 8265, 9260 (aka Killer 1550) and Qualcomm QCA6174/A (aka Killer 1535) and a single antenna QCA card whose name I forget but is found in quite a few cheap Acer laptops. You also have Broadcom clients like the Galaxy S7 which end up dropping to a single antenna mode when MU is enabled.

 

[doczenith1]

Ok here goes. I tested using a Pixel XL and a Pixel 2 XL with my 86U on 384.3. Phones were sitting 1 inch apart on the coffee table about 12 feet and one wall away from the router. Each had a 866 link speed. I ran the test using the web based DSLReports speed test. I ran 5 tests and took the average of the best 4 values. Results are DL/UL in Mbps.

Testing phones one at a time to set a baseline.
Pixel XL 582/479
Pixel 2 XL 536/571

Testing phones at the same time with MU-MIMO disabled
Pixel XL 410/329
Pixel 2 XL 274/410

After this test I "forgot" the wifi connection on each phone. I enabled MU-MIMO and rebooted the router.

Testing phones at the same time with MU-MIMO enabled
Pixel XL 264/145
Pixel 2 XL 272/468

The values for both tests when testing both phones at the same time were all over the place. Looking at the results it appears that total throughput is higher with MU-MIMO disabled in my environment and configuration.

I do have a non-MU-MIMO 2x2 Moto X Pure Ed that I could add to the mix if there is value in testing all three together. Just let me know.

In thinking more about this topic, I wonder, is it possible for router makers to detect which clients support MU-MIMO and provide that data in the devices list?

On the System Log>Wireless Log page both of my phones show an "M" flag and per the legend M=MU Beamforming.

 

[avtella]

Ok here goes. I tested using a Pixel XL and a Pixel 2 XL with my 86U on 384.3. Phones were sitting 1 inch apart on the coffee table about 12 feet and one wall away from the router. Each had a 866 link speed. I ran the test using the web based DSLReports speed test. I ran 5 tests and took the average of the best 4 values. Results are DL/UL in Mbps.

Testing phones one at a time to set a baseline.
Pixel XL 582/479
Pixel 2 XL 536/571

Testing phones at the same time with MU-MIMO disabled
Pixel XL 410/329
Pixel 2 XL 274/410

After this test I "forgot" the wifi connection on each phone. I enabled MU-MIMO and rebooted the router.

Testing phones at the same time with MU-MIMO enabled
Pixel XL 264/145
Pixel 2 XL 272/468

The values for both tests when testing both phones at the same time were all over the place. Looking at the results it appears that total throughput is higher with MU-MIMO disabled in my environment and configuration.

I do have a non-MU-MIMO 2x2 Moto X Pure Ed that I could add to the mix if there is value in testing all three together. Just let me know.



On the System Log>Wireless Log page both of my phones show an "M" flag and per the legend M=MU Beamforming.

I read somewhere that it's less advantageous when devices are too close in proximity. Can you try like 5ft apart or something? I expected an improvement with the 4366E on 86U.

 

[thiggins]

Avtella is correct. You need some distance between the devices so that proper beamforming can be done. Can be a few feet, but not inches.

In my tests, I didn't see much difference between having the devices a foot or so apart and one in front and one in back or to the side of the router.