Beamforming / Mu-MIMO / Airtime Fairness

[leerees]

EDITED to clarify:
At the RISK of backing up a Bit...
I feel this statement may be one of the most beneficial to understand & I think it touches on some very misunderstood aspects regarding router radio channels, Access Points, & Mesh Networks.
My question stems from me recently enabling AiMesh & using the Android app Wifi Analyzer on a phone.
I was somewhat disappointed to see, despite me turning down the power (which I could only do on the primary router) & trying to separate the physical distance of my two RT-AC68U (One=Node of course)...
The Parabolic Channel Curves of the primary router & node directly overlap one-another.
It is my "limited understanding" that basically any waveform overlap is "possibly" INTERFERENCE & "maybe" only the difference between the two curves is USABLE bandwidth?
Is this logic, Correct?
If so... to maximize WiFi throughput... I'd be better off putting the node in AP mode, & ensuring my two RT-AC68U's did not overlap with each-other.
Thus WiFi clients connected to 2-different RT-AC68U devices BUT it would most likely leave more bandwidth available for network control.
+I feel the other MISUNDERSTOOD point (especially with Mesh) is...
Wouldn't we typically be better off lowering our (Router/AP) power as much as possible (yet still maintain desired WiFi connections) to each AP.
Especially if we could get our neighbors to do so... but not always possible.
I think people forget the idea with Mesh should be, More Nodes @Less Power.
Simply deploying More Nodes & Blasting at Max Power will further pollute nearby Radio waves.

But with MESH isn't it still best to eliminate any overlap (even between a Mesh router & Node)... so far with my AiMesh it doesn't seem to behave as such.
As Tech9 mentioned earlier, channel separation can also be done by narrowing the channels.
But the trade off is less throughput or bandwidth available for a narrower wifi channel.

Or am I totally missing something & with Mesh it's totally OK to have the two different devices overlap each-other???

You can adjust the power levels with Asus.

You can't put nodes on different channels with AI Mesh. Maybe Merlin firmware might allow this.

 

[thiggins]

@capncybo If mesh nodes connect via Wi-Fi and they have only two radios (one each 2.4 and 5 GHz), then at least one radio in each node must be on the same channel as the radio in other nodes it forms a mesh with. Otherwise they can't pass packets through the mesh to client devices. This is known as the "backhaul" connection.

When Ethernet is used as the mesh backhaul, both radios in each mesh node can be freely set. Best practice in Wi-Fi network design limits transmit power of each AP's radio and assigns adjacent APs to non-overlapping channels. This lets each AP provide its own bandwidth to connected clients without competing for airtime with other APs.

Yes, Ethernet is always preferred over Wi-Fi as backhaul since it has higher and more stable bandwidth.

 

[dlandiss]

If mesh nodes connect via Wi-Fi and they have only two radios (one each 2.4 and 5 GHz), then at least one radio in each node must be on the same channel as the radio in other nodes it forms a mesh with.

Do I recall correctly this "same channel" choice is actually preferable to using a nearby channel because the 802.11 protocols allow same-channel radios to implement multiplex schemes that avoid mutual interference, while adjacent-channel radios cannot do the same?

In other words, you are better off sticking with a 1-6-11 scheme on 2.4GHz because channel 5 (let's say) could interfere with channel 6 more than another channel 6 radio would?

Thanks!

 

[thiggins]

Devices must be set to the same channel to communicate. One channel over doesn't do it.
Now if you are asking whether channels other than 1,6,11 can be used for APs connected via Ethernet, I refer you to the "Channels" section of this oldie.

How To Fix Your Wireless Network - Part 1 - SmallNetBuilder

http://forums.smallnetbuilder.com/forumdisplay.php?f=18

www.smallnetbuilder.com

 

[capncybo]

You can adjust the power levels with Asus.

You can't put nodes on different channels with AI Mesh. Maybe Merlin firmware might allow this.

Yes (thank you) I did know that & have already lowered the power on the Mesh-Router... What I meant was, I expected the Node to follow suite and basically mirror the power settings of the router.
I'm not @home right now, but (from memory) when using my phone & wifi analyzer... I think the Node channel amplitude did not drop & mirror the router settings.

 

[capncybo]

@capncybo If mesh nodes connect via Wi-Fi and they have only two radios (one each 2.4 and 5 GHz), then at least one radio in each node must be on the same channel as the radio in other nodes it forms a mesh with. Otherwise they can't pass packets through the mesh to client devices. This is known as the "backhaul" connection.

When Ethernet is used as the mesh backhaul, both radios in each mesh node can be freely set. Best practice in Wi-Fi network design limits transmit power of each AP's radio and assigns adjacent APs to non-overlapping channels. This lets each AP provide its own bandwidth to connected clients without competing for airtime with other APs.

Yes, Ethernet is always preferred over Wi-Fi as backhaul since it has higher and more stable bandwidth.

@thiggins I appreciate the reply.
I did know about the backhaul using bandwidth, but I never visualized what that might look like from a wireless analyzer perspective.
So according to your info...
If I pull out my 200foot Ethernet cable & string it down the hallway (between Router & Node) & then: [check, set, & verify] the Router & Node are using Ethernet(wired) backhaul.
My two overlapping wifi SSID control channels should actually separate & not overlap one another.
Correct?

 

[L&LD]

Not @thiggins, but...

No.

 

[capncybo]

Not @thiggins, but...

No.

Okay then... from PURELY a WiFi throughput standpoint...
It's probably still better to try & separate channels between router & AP.
(But as stated above Nodes stay on the same channel.)
Is the main & perhaps only advantage of AiMesh that network managment is centralized & it greatly simplifies your network topology??

Or does AiMesh also implement some additional & perhaps propitiatory tricks to improve performance?

To rephrase the question...
When using a Wired Ethernet Backhaul... can AiMesh possibly equal or Better the performance of the more traditional Router-AP-AP topology?

 

[L&LD]

Yes, depending on the situation and purely from a theoretical point, it is better to separate control channels for a specific (single) network. However, looking at it holistically, with many AP's surrounding most networks, this doesn't pan out in practice (interference from other networks is just as detrimental, if not more so).

AiMesh does simply to create a repeater (when using wireless backhaul), yes.

Unless you're using a true tri-radio or better router, AiMesh cannot use any 'tricks' for your network.

When using wired Ethernet backhaul, and those main routers and nodes are put as far apart as necessary, the performance is identical to using a single main router that is always in close proximity. The benefits are now mostly on the management side vs. a 'dumb' AP. With more features and capabilities coming with each new firmware released.

 

[Tech9]

When using a Wired Ethernet Backhaul... can AiMesh possibly equal or Better the performance of the more traditional Router-AP-AP topology?

I did test that - the performance of Router-AP is better. The roaming is exactly the same, AiMesh doesn't improve anything. The same radios support the same roaming assist technologies. Router-AP allows power control and use of Roaming Assistant in Asuswrt, when it works. I found on some firmware versions it does nothing. AiMesh with wireless backhaul is Router + Repeater. AiMesh with wired backhaul is Router + AP with central management UI in exchange of less AP control. It's made to be easy for the consumers so they can go buy another Asus router, not better.

 

[Phantomski]

MU MIMO: Again, crucial to allow your router to communicate with multiple streams simultaneously. Disabling it will impact the performance of your WiFi. If you manage to find an IOT device that isn't compatible with this (i haven't) then you need to put that IOT device on eBay and be done with it.

I find this mildly amusing. Most of the IoT these days is still barely 2.4GHz 802.11g, as the majority of these cheap low energy chips they use still are. For me personally, out of 50+ devices I have, I’d have to eBay 45 of them with no decent (and ac or ax 5G for that matter) replacement available.

Not a good advice. Sorry.

 

[capncybo]

I did test that - the performance of Router-AP is better. The roaming is exactly the same, AiMesh doesn't improve anything. The same radios support the same roaming assist technologies. Router-AP allows power control and use of Roaming Assistant in Asuswrt, when it works. I found on some firmware versions it does nothing. AiMesh with wireless backhaul is Router + Repeater. AiMesh with wired backhaul is Router + AP with central management UI in exchange of less AP control. It's made to be easy for the consumers so they can go buy another Asus router, not better.

Ahhh I was hoping for your opinion since posting my initial question (Post#59), as I feel you have demonstrated a very thorough understanding of How Most of this Magic Works! Your above statement sums up my current opinion nicely, so Thank-you. + I have since educated myself further regarding WiFi Analyzer Signal Overlap. What I'll share with others is... apparently Direct Overlapping (or what I'll call stacked) signal is called Co-Channel interference. However this form of overlap is not actually as Horrific as it appears. What it represents is Wifi Data that is actually readable, & sometimes waiting in a Que. So with two linked routers or AP it would manifest as congestion, increased latency or delay.

 

[dlandiss]

Devices must be set to the same channel to communicate. One channel over doesn't do it.
Now if you are asking whether channels other than 1,6,11 can be used for APs connected via Ethernet, I refer you to the "Channels" section of this oldie.

How To Fix Your Wireless Network - Part 1 - SmallNetBuilder

http://forums.smallnetbuilder.com/forumdisplay.php?f=18

www.smallnetbuilder.com

Thanks, Tim. I know that for a repeating function the channels have to be the same at both ends, else how could they have 2-way comms?

But some modern mesh networks seem to put the remote nodes on channels others than 1-6-11. I also understand why that is likely to cause trouble -- I taught electronic communications for 34 years we even had spectrum analyzer experiments in our class. My concern is that nobody seems to be calling out the mesh manufacturers, and the customers don't even know it is happening.

Below is a clip prepared by MetaGeek that clearly shows the spillover between channels.

 

[thiggins]

But some modern mesh networks seem to put the remote nodes on channels others than 1-6-11.

It's not just mesh systems. I've seen standard routers use auto channel selection and end up on 4 or 2.4 GHz channels other then 1-6-11.

 

[thiggins]

apparently Direct Overlapping (or what I'll call stacked) signal is called Co-Channel interference. However this form of overlap is not actually as Horrific as it appears. What it represents is Wifi Data that is actually readable, & sometimes waiting in a Que. So with two linked routers or AP it would manifest as congestion, increased latency or delay.

There would only be congestion if the channel were heavily used or oversubscribed. This video might help explain.

 

[dlandiss]

It's not just mesh systems. I've seen standard routers use auto channel selection and end up on 4 or 2.4 GHz channels other then 1-6-11.

Yep. It appears that the "auto" choice is made based only on control-channel traffic ("amplitude"), ignoring proximity to neighboring channels ("frequency").

 

[Tech9]

Your above statement sums up my current opinion nicely, so Thank-you.

There was one more opinion in this thread, but I've decided to delete it. One of the forum members here not only doesn't understand how the things work, but the entire story presented doesn't make sense from both technical and practical points of view. From my experience with multi-AP networks, I see too much sugar coating. I may get diabetes, if I continue posting my opinion. I'm sorry, @capncybo. Test what you have there and see what it does.

But some modern mesh networks seem to put the remote nodes on channels others than 1-6-11.

Of course. 1-6-11 is valid only when you control all the APs. How many home router owners control neighbor's APs? This rule doesn't apply for home routers and very often folks end up using the worst channels possible with the least available bandwidth. Run your modern router on Auto and see what channel it uses the most. The selection comes from wireless drivers, not the device manufacturer. It will use the best available bandwidth channel.

 

[dlandiss]

Of course. 1-6-11 is valid only when you control all the APs. How many home router owners control neighbor's APs? This rule doesn't apply for home routers and very often folks end up using the worst channels possible with the least available bandwidth. Run your modern router on Auto and see what channel it uses the most. The selection comes from wireless drivers, not the device manufacturer. It will use the best available bandwidth channel.

The problem being that "best" is determined by an algorithm that does not take adjacent-channel interference into account. Here is what I see in my office:

 

[Tech9]

It does. Included in Asuswrt-Merlin Wi-Fi Analyzer shows what Broadcom drivers see. There is interference information as well, as the router sees it. In my Wi-Fi environment Ch.4 on 2.4GHz offers the best throughput. The adjacent channels are strong, but mostly inactive. This is what most folks get wrong when looking for “available” channels - a strong channel in phone Wi-Fi Analyzer app may be beacons only from a close router. The same channel or close to it may have the best bandwidth available. Today’s routers are more intelligent in channel selection any take into account multiple factors.

 

[dlandiss]

The adjacent channels are strong, but mostly inactive.

Fortunately, as you can see from that spectrum clip, most of my interference candidates are below -70 RSSI. The maverick on Ch 3 is the main exception. In any case I have gotten splendid results with my router on Ch 1 and its AP on Ch 11.

Interference on 5GHz is not an issue, but it doesn't propagate well in my 119-year-old house.