Does a router always bond channels as listed in this table?

[sittingduck]

Trying to understand how channel bonding works on a router. If I have channel 48 as my control channel and router set to autonegotiate 20/40/80 bandwidth, will the router always bond channels per this table - eg. always bonding 44+48 for a 40 mhz channel, or could it also bond 48 and 52 sometimes instead? Thanks.

 

[ColinTaylor]

Trying to understand how channel bonding works on a router. If I have channel 48 as my control channel and router set to autonegotiate 20/40/80 bandwidth, will the router always bond channels per this table - eg. always bonding 44+48 for a 40 mhz channel, or could it also bond 48 and 52 sometimes instead? Thanks.

It always combines the channels in the way shown in the wiki. So channel 48+52 is not a valid combination for 40MHz.

 

[tgl]

@ColinTaylor is right, but it might also be worth noting that this applies for the 5GHz band. 6GHz has similar rules about ganging channels in a specific pattern, but 2.4GHz does not --- if you choose a 40MHz channel there, you just spill over onto adjacent channels in both directions. That's one reason why it's generally a bad idea to choose 40MHz in the 2.4GHz band. The channel ganging patterns in the upper bands were chosen to ensure that channels don't overlap unnecessarily, but the standards for 2.4GHz are older and less well written.

 

[Tech9]

you just spill over onto adjacent channels in both directions

One can still choose upper or lower channel bonding with control channel somewhere around the middle of the band.

 

[tgl]

One can still choose upper or lower channel bonding with control channel somewhere around the middle of the band.

Right. My point is that wide channels in 2.4GHz are the wild west: everybody can choose their own random frequency range, and pretty much guarantee that they're overlapping with all their neighbors not just some neighbors. This doesn't matter if you're in the boonies with no visible neighboring SSIDs, but for city living it's one of the reasons why 2.4GHz is on the edge of uselessness.

 

[sfx2000]

Much of this is driver implementation...

If your primary/control channel is 48, makes sense that 44 is the preferred choice in most markets as 44 is not a DFS channel, where 52 would be a DFS selection.

The better question - what if 44 is the control channel, and one is choosing 40 MHz channels for 5GHz - there it gets interesting with implementation as it could go up or down, and this would apply for 40/80 MHz

Here's a fun question - what if your 5GHz Control Channel is 165 - what are the permitted secondary channels and permitted bandwidths for 11ac/11ax/1be?

 

[tgl]

The better question - what if 44 is the control channel, and one is choosing 40 MHz channels for 5GHz - there it gets interesting with implementation as it could go up or down

I don't believe that's correct. AFAIK, the 5GHz/6GHz channel bonding hierarchy described in that wikipedia page is required by the wifi standards.

 

[sfx2000]

Right. My point is that wide channels in 2.4GHz are the wild west: everybody can choose their own random frequency range, and pretty much guarantee that they're overlapping with all their neighbors not just some neighbors.

2.4 actually has rules here about the secondaries, as it will always be plus or minus 4 and stay within the band... e.g. if the primary/control is 6, the secondary will either be 10 or 2 with no hard reason for either, and if the control is 3, the only possible one would be 7...

The old 1/6/11 rule was valid for 11g, but it's kind of not these days with 11ax/be, and to some degree even with 11n...

 

[ColinTaylor]

I don't believe that's correct. AFAIK, the 5GHz/6GHz channel bonding hierarchy described in that wikipedia page is required by the wifi standards.

Agreed.

 

[sfx2000]

I don't believe that's correct. AFAIK, the 5GHz/6GHz channel bonding hierarchy described in that wikipedia page is required by the wifi standards.

Go back and read it again - 44 can be up or down - there is no standard defined there...

wikipedia is not a normative reference for anything... best would be community consensus perhaps...

 

[Tech9]

I don't remember seeing any Wi-Fi equipment allowing control channel 44 @40MHz bandwidth with anything else than 48. This is commonly called channel 46 and can have control channel 44 or 48. This is my EU gateway with allowed AP channel configurations:

 

[ColinTaylor]

44 can be up or down - there is no standard defined there...

We were talking about channel 48+44 @40MHz. It can't be 48+52 (which was the OP's question). In your example where channel 44 is the primary, @40MHz the secondary is 48. It can't "go up or down". It is a standard.