How to determine the best channel?

Muis

Occasional Visitor
I wonder how to determine the best wi-fi channel. I used a wifi-analyzer, and it shows only a small amount of other routers on channel 1, but those ones have a relatively strong signal. And it shows a big amount amount of routers on channel 11, but those one ones have a weak signal.

So hypotetically speaking, is it better to pick the channel with the least other routers or the channel with the least nearby ones?
 

thiggins

Mr. Easy
Staff member
Neither. It's best to choose the channel with the least sustained activity (torrents, large downloads). Unfortunately, there are no free options I know of that provide this information.

Metageek's inSSIDer with a Metageek Plus subscription can do this. There is a free trial.

Unfortunately, the subscription is $200/yr, with no monthly option. So the company clearly is not focused on home users.
 

Muis

Occasional Visitor
Neither. It's best to choose the channel with the least sustained activity (torrents, large downloads).

I understand that if you had more information, like sustained activity, you could make a much better choice. But let's just assume for now that the activity on all networks is the same, would it generally be better to pick the channel with the least other routers, or the channel with the least nearby ones?
 

L&LD

Part of the Furniture
Simply test each Control Channel, keep good notes, use the one with the least compromises in your current environment.

 

OzarkEdge

Part of the Furniture
Simply test each Control Channel, keep good notes, use the one with the least compromises in your current environment.

... until further notice from your users that the network is slow! Then start over... like after Christmas when the neighbors install their shiny new misconfigured multi-node mesh systems. :)

OE
 

Muis

Occasional Visitor
Simply test each Control Channel, keep good notes, use the one with the least compromises in your current environment.

But what is a good way to test these channels? Is it just a matter of nothing down each of their dbm values? Does this reflect the interference/congestion accurately? Or do I need to run filetransfers to measure the maximal speed of each channel?
 

L&LD

Part of the Furniture
Don't bother about the 'numbers'. Maximum speeds are not your primary goals, consistency and lowest latency are more important. dBm values are worthless unless you're designing your own circuits (end to end).

Change the Control Channel to anything but Auto. Reboot the router. Verify that the Control Channel you picked, sticks. Then simply use your network normally.

Take notes on important things. Like fringe areas with certain channels. And/or with certain client devices. How responsive the network feels (lower latency is always better than marginally faster speeds) for both web browsing and internal, LAN, workloads (like accessing a NAS, printer, scanner, etc.). And how stable (and responsive) the network is over a course of the day (when the neighbors come home, and the streaming/gaming starts).

Anything else you keep notes on is incidental and not important, as it won't help you get a faster networking experience any sooner or more reliably either.

The WiFi apps are great to see 'stats', but they don't do anything to improve your network (except by chance/luck and a lot of wasted time).

Remember when you're testing that you're not relying on a single client device to do all your tests with (this isn't a race). Use all the clients you have in all the areas that they are normally used and use them normally (speedtest.net and fast.com are not normal uses). These are the notes you take and compare across different channels used.

There are only so many Control Channels to 'test'. And these can be done very quickly to see what is the best channel 'now'. Once you have preliminarily decided on the best candidate channel to use, be sure you leave it for at least 24Hrs or more before you give up on it. This allows the surrounding APs (which are most likely on Auto) to move off your selected channel.

Did you read the following too, from the link, I provided above?

Control Channel Setup 2021

Control Channel Setup (more)
 

CaptainSTX

Part of the Furniture
Trying the best 2.4Ghz channel in a crowded environment is never ending task as some of your neighbors will have their channel selection set to automatic so their router will randomly-change channels so the tests you do today to find the least busy channel won't be applicable tomorrow. This is even a bigger problem if they have selected to use wide channels.
 

L&LD

Part of the Furniture
Good WiFi in a crowded environment is a never-ending task. Yes.
 

Dark910

Occasional Visitor
I usually go about testing only 1,6, or 11 to prevent interference as much as possible. Then I make my determination based off how the AP/router behaves on each of those three channels, and how my clients also react. Not all clients and not all routers/AP will have the same transmit power per channel. Sometimes it will be noticeably more on one than the other. Usually 6 is higher than 1 or 11, but not always. It's not a big deal whether there are networks already unless they're busy(unless it's 2.4 GHz and there's way too many SSID beacons taking up airtime in an extreme situation), and it's possible they'll just move anyway if you find one channel is better. It's still worth testing the channel to make sure it's not a complete trash town of interference though. For 5 GHz, the same sort of practice. Sometimes the low-band(in the US) can be less transmit power than the upper band channels on some devices, so low-band tends to be better for trickier areas and upper band for more direct pathways as I think it's better, but falls off more quickly. If you have weak client devices, that's also something to take into consideration, as they usually don't have a lot of power and the higher the frequency between channels(specifically mentioning 5 GHz) may cause issues, especially if you need upload.

In regards to power differences between channels, it's more often the case on routers that aren't top of the line. You can gain a few dB just by using channels your device works best on, then just let the neighboring devices figure it out and move.

Also wanted to note that just because a channel shows less networks, doesn't mean that there necessarily are less networks when you do a scan. If a channel is congested, you probably won't see any networks below a certain signal range and only the higher strength ones while the channel is being loaded up. Ideally though you probably want the neighboring networks to be about -75 dBm in strength or less. From my understanding, there is a pretty long wave of noise produced outside the primary width that affects most of a wireless band, but it's very low strength varying by device and the neighboring device needs to be fairly close/high strength to get hit by it, and in that case I'd suggest you just stay on the same channel to promote co-channel interference management, but you may not run into that unless you're extremely close to some other very active network(of which may not work if the other device just moves away anyway).
 
Last edited:

Sign Up For SNBForums Daily Digest

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