Starting to despair

[TheLyppardMan]

I've currently set it to auto as others have advised above. I'll see how it goes for the next day or so.

This morning, I tried channel 11 again. The problem happened for a while this morning and then it was OK. This afternoon, just after 2.00 PM, I switched on 20/40 MHz auto mode and it's been running fine on that setting so far (using channel 7 at 20 MHz).

These are the results from earlier in the day. The ones at the bottom of the list are the earlier results:

And these are the results from this afternoon:

 

[L&LD]

You're assuming that Auto is an intelligent setting. It's not.

It uses primative algorithms to make it seem like it is doing something. And while it may guess correctly randomly, it excels primarily at giving a variable network performance (almost guaranteed, for most).

It isn't taking into account channel utilization, nor non-WiFi sources of interference. And in many cases, it chooses a control channel that is just obviously wrong too.

But, it may work for you. (I don't waste time hoping it might).

 

[drinkingbird]

Auto is a good way to have your network performance be variable.

Test each channel for an extended period (24 hours +), keep good notes, and use the best one.

Control Channel Setup 2021

Reset Mini Guide + Control Channel Setup Details

Control Channel Setup (more)

And then do that every single day and pick the best channel for that particular day. Actually probably want to do it multiple times a day.

 

[drinkingbird]

I've currently set it to auto as others have advised above. I'll see how it goes for the next day or so.

This morning, I tried channel 11 again. The problem happened for a while this morning and then it was OK. This afternoon, just after 2.00 PM, I switched on 20/40 MHz auto mode and it's been running fine on that setting so far (using channel 7 at 20 MHz).

These are the results from earlier in the day. The ones at the bottom of the list are the earlier results:



And these are the results from this afternoon:

Why turn on 20/40? Why keep trying 11? I don't get what you're trying to do other than create hassle and instability for yourself.

 

[drinkingbird]

You're assuming that Auto is an intelligent setting. It's not.

It uses primative algorithms to make it seem like it is doing something. And while it may guess correctly randomly, it excels primarily at giving a variable network performance (almost guaranteed, for most).

It isn't taking into account channel utilization, nor non-WiFi sources of interference. And in many cases, it chooses a control channel that is just obviously wrong too.

But, it may work for you. (I don't waste time hoping it might).

Personally in my environment I had to give up on hardcoding, I was having to find and change to a new channel every few days to a week. Auto has been working perfectly fine on both my Asus and my Ubiquiti for a couple years now, and I've done testing and confirmed the channels they are picking are at or near the top performance wise. The ubiquiti is a bit more advanced in its scanning but both are working and doing what they should.

If you're in a rural area then hardcoding makes sense, but if you have even a few neighbors nearby, just doesn't cut it from my experience. Yes you become part of the musical chairs game but better to play than to sit out and get nothing.

 

[L&LD]

And then do that every single day and pick the best channel for that particular day. Actually probably want to do it multiple times a day.

No, you don't do that. Read the links above. Notes help here (testing usually lasts a week or more).

Environments don't change that often. There is no ultimate/single Control Channel to choose. There is only the best one with the least compromises for a particular environment with the mix of client devices currently in use.

We're not trying to mimic the poor choices Auto makes, after all.

 

[tnpapa]

I am in the same boat. I have to use Auto. I have dozens of nearby networks that are all constantly changing what channel they are on. If I just hardcode a channel it may be good for a day, or maybe an hour. But then all the other nearby routers decide to randomly change channel and they land where I am, and my signal goes to poop. To stay manual I would have to open my spectrum analyzer and see what is open and change channels. Too much work to do everyday. So I let the router figure it out no matter how flawed its algo may be, its easier than me doing it.

 

[drinkingbird]

No, you don't do that. Read the links above. Notes help here (testing usually lasts a week or more).

Environments don't change that often. There is no ultimate/single Control Channel to choose. There is only the best one with the least compromises for a particular environment with the mix of client devices currently in use.

We're not trying to mimic the poor choices Auto makes, after all.

Residential suburban or urban environments DO change that often.

If you have total control over the environment with no neighboring networks (or very weak ones) like in a commercial setting or rural area, then the suggestion/recommendation is totally different. But for the average residential user, Auto is likely to result in the best performance over time. Yes, it may vary still, but hopefully avoid those days when you have barely any bandwidth due to saturation. In my case anyway, that's been exactly the case, where I'd drop well under 10M on days when I was hardcoded, have never had that happen since switching to auto (as much as it pained me to switch).

It also pains me to see non 1/6/11 channels, but that's just the reality these days.

 

[L&LD]

No, they don't. At least not over a 24-hour or more period.

And even if they do, it is only temporary.

If you want the most consistent network, Auto CCs aren't the way.

 

[tnpapa]

No, they don't. At least not over a 24-hour or more period.

And even if they do, it is only temporary.

If you want the most consistent network, Auto CCs aren't the way.

I am sorry but you are dead wrong. I watch my neighbors wifi change channels constantly. sometimes while I am monitoring the spectrum with wifi explorer I see them all jumping around. One changes channels, and that triggers the algo to fire on another neighbors router, so they change channels. It never stops.

Sorry L&LD but I just don't subscribe to your beliefs about how anything in the ASUS routers work. I have read many threads of yours where others have pointed out the flaws in your logic.

 

[TheLyppardMan]

Wow, quite a lot of conflicting views here. I'm going to leave my settings on auto for now and see what happens over the next few days. I'll report back when I have further information to hand.

 

[Deleted member 84281]

Environments don't change that often.

Yes they do. Routers set to auto change channels all the time. If you find a clean fixed channel use it but be aware any neighboring router can join you there at any second. This is why wifi analizers are useless. What you see one minute can change the next. Often AP will find each other because being on the same channel is better then cross splatter from another. 2.4Ghz should just be avoided period unless you live in the forest. It's simply not needed today and should be disabled.

 

[Morris]

As you may have seen in my other thread, I had a big problem with the 2.4 GHz Wi-Fi band on my RT-AX88U Pro, so I sent my router back to Amazon and I'm currently waiting for a refund. The router was on special offer at Amazon earlier this week, so I took a chance and ordered a new one. Guess what - it's exhibiting similar behaviour to the one I have just sent back. Can anyone offer a possible explanation as to what is going on here and what I should do? I've spend many hours configuring and then wall-mounting these routers and I'm getting fed up with this now. Please help if you can.

I had some odd behavior for devices on my 2.4-Ghz band. Then I found this:

How to improve compatibility of IoT devices with WiFi 6 (AX) Router

[Wireless Router] How to improve compatibility of IoT device with ASUS WiFi 6(802.11 ax) router? | Official Support | ASUS Global

All 2.4-Ghz devices have been happy ever after.

 

[L&LD]

I am sorry but you are dead wrong. I watch my neighbors wifi change channels constantly. sometimes while I am monitoring the spectrum with wifi explorer I see them all jumping around. One changes channels, and that triggers the algo to fire on another neighbors router, so they change channels. It never stops.

Sorry L&LD but I just don't subscribe to your beliefs about how anything in the ASUS routers work. I have read many threads of yours where others have pointed out the flaws in your logic.

No, I'm not wrong. You're assuming that the rest of the neighborhood control channels on Auto have an impact on your carefully chosen manual control channel. Think again.

What the process does is actually find the best control channel for the environment. Via observation.

What your logic states is that other routers' Auto settings are superior.

This is not what I observe in real life.

None of my networks (mine and the customers I set them up for) have needed changing for the same router/client devices in almost a decade.

Meanwhile, Auto CC changes multiple times a day.

My logic is not flawed, in anything I've written on these forums. Review yours.

Yes they do. Routers set to auto change channels all the time. If you find a clean fixed channel use it but be aware any neighboring router can join you there at any second. This is why wifi analizers are useless. What you see one minute can change the next. Often AP will find each other because being on the same channel is better then cross splatter from another. 2.4Ghz should just be avoided period unless you live in the forest. It's simply not needed today and should be disabled.

To reiterate and expand on the response above in this post, I'm not suggesting choosing the best channel 'now'. You keep missing this point.

All the reasons you state after that I can (mostly) agree with. But that doesn't mean a new control channel needs to be chosen, right now. This is what leads to varying network performance, and that is more annoying to most people than having the possible highest speeds (usually, just a few Mbps faster) 'sometimes'.


To both: If Auto really worked as it should, we wouldn't be having this discussion. But it doesn't.

Believe what you want. Anyone reading along now or in the future has the choice to test both methods in their own environments and see how it works for them.

 

[tnpapa]

You can rant all you want, but my real world observations of my neighbors wifi over the years is not deniable. With over 24 networks all within range of each other and various brands of routers there is NEVER any one channel that is more stable than the other for more than a few hours. I can have a channel almost to myself and get terrific speeds, only an hour later to have multiple devices drop off my network because there are now 15 other routers on the same channel and they are wiping out my connection. Multiple routers will jump channels together or within a short time of each other. Sorry but the real world is not what you imagine. It is exactly the flawed or bad algos of all these routers that makes my picking any one channel futile.

Your logic is without proof, other than what you observe, which is not what I observe. You can't claim superiority or all knowing. I am not going to argue this anymore because you can't change the physics of my RF reality. I am also blocking you now because I don't even want to continue this any further.

 

[Morris]

All, there is a clastic problem with WiFi auto channel where one AP changes channel as it detects congestion on the channel it is using. It changes to a channel that seems clear yet when it changes other APs that are using this channel now detect the channel as congested and change channels creating a wave that looks much like the wave fans do in a sports stadium. This can go on forever. Every manufacture has there own way of defending from this situation yet a common approach is to avoid changing channels more frequently than X minutes. This is to avoid clients having a difficult time staying connected. A fixed channel selection, even one with some congestion can be more stable in situations like this.

 

[DJones]

Amusingly I’m the bad neighbour type in this situation just using ASUS channelhog hogging a bunch of channels with my main router and mesh. I live rural and the few neighbours wifi that I can detect seem to keep things auto so I hardcode to non overlapping and push them to random channels that causes them congestion. If that’s efficient or not for me doesn’t really bother me doesn’t seems to cause me much trouble with throughput.

 

[L&LD]

@tnpapa, your points don't validate your wrong conclusions from the observations you see. They validate mine even more.

And of course, sticking your head in the sand is the right thing to do when someone has a different opinion and backs them up (as others have too, see above posts).

 

[sfx2000]

I still feel the 40mhz rule applies - differing opinions don't have to be kept private.

Sometimes theoretical testing and real world are vastly different (more than sometimes in my experience).

Fair enough - the 1/6/11 alignments were based on 11b with 22.5MHz channels...

With OFDM using 20MHz channels, we can get better utilization with 1,4,8,11 (and in some countries, 13)

11n co-channel interfere rejection was planned around wide-channel usage - and that includes the CCA checks for the secondary channel...

Bandwidth is not a precious thing to be conserved - rather it is something to be used or it's lost...

There's a couple of reasons to do 20MHz channels in 2.4...

1) if you're apple centric - no Apple device will do wide channels in 2.4 - used to be that they would actually send out fat channel intolerant on probes, but this has changed over time, but Mac/iPad/iPhone devices do 20MHz in 2.4GHz...

2) Wide Channels have a 3 dB hit on the link budget - this is true, and this will impact speed over range, but even there, once you get to a certain range, it doesn't really matter...

Oh, BTW - once you get to a certain level of neighboring AP's, it's all just noise floor - doesn't matter whether wide or narrow at that point, so again, might as just well go wide...

Being "neighbor friendly" - that doesn't really hold water, mostly because your neighbor gives zero-f**ks about your wifi, so why bother worrying about them?

I say try wide channels - worst case, no change in performance, best case is you'll increase performance - it won't double, but it will improve...

Leave the chips on the table, or use them - that's my advice...

 

[drinkingbird]

With OFDM using 20MHz channels, we can get better utilization with 1,4,8,11 (and in some countries, 13)

I've also seen 1/3/7/11. In reality, unless you control the entire environment, none of the "rules" apply. I'd love to coordinate my neighborhood's channels and limit everyone to 20 mhz but pretty sure I'd get more than a few doors slammed in my face.

Given the saturation and challenges, I still think 40mhz never should have been introduced and certainly shouldn't be a default (even if the router supposedly checks before enabling it). At the very least, it should not kick in until a device capable of using it is present.

We are not sacrificing any bandwidth here, it is fully utilized.

I've seen 40mhz get worse performance than 20mhz (much like you can see worse performance with 160 vs 80 vs 40 vs 20), so turning it on and "letting the chips fall" is isn't necessarily the case, especially since you're pretty much guaranteed to be introducing more noise/interference/contention into your network.