2.4GHz Channels?

[Synomenon]

I'm setting up a new router. Before setting the channel for the 2.4GHz radio, I loaded up an Android app. from Amped Wireless. The app. scans the area for other wifi networks and tells you which channels are least congested. It says channel 8 is the least congested. Should I switch my router's 2.4GHz radio to channel 8 or leave it on channel 1, 6, or 11?

 

[stevech]

use 1, 6 or 11. These non-overlap. Channel 8, for example, gets competition for clear-air time from both 6 and 11.

The best choice is the one that has no/little use of video/Netflix via WiFi or big/continuous file transfers.

Unfortunately, the usual WiFi tools show you what SSIDs are nearby and their signal strength. This tells you nothing much useful about how busy those SSIDs are.

An unused WiFi access point/router still sends beacons. That's what your WiFi tool relies upon to make a list by channel. The beacons alone use a tiny fraction of a percent of the available channel capacity.

A good indicator is to run ping on your client PC: ping -t 192.168.1.1 (or whatever your WiFi router/AP address is). Let this run for a time during the evening Netflix times. See if the latency numbers tend to AVERAGE higher at some times on some channels (1, 6 or 11).

Or for the geekly included, automate the above using http://www.pingplotter.com/ which is free if you work your way through the attempts to get you to pay.

 

[Mark Uhde]

I'm setting up a new router. Before setting the channel for the 2.4GHz radio, I loaded up an Android app. from Amped Wireless. The app. scans the area for other wifi networks and tells you which channels are least congested. It says channel 8 is the least congested. Should I switch my router's 2.4GHz radio to channel 8 or leave it on channel 1, 6, or 11?

You don't really know that 1,6,11 are the non-overlapping channels in your area. In general, the previous post is correct, but please download inSSIDer if on Windows or iStumbler if on Mac and post a list of all the SSID's in your area, their signal strengths, and their channel numbers and it will be possible to give you a much better (though not perfect without spectrum analysis) channel number to use.

 

[stevech]

You don't really know that 1,6,11 are the non-overlapping channels in your area. In general, the previous post is correct, but please download inSSIDer if on Windows or iStumbler if on Mac and post a list of all the SSID's in your area, their signal strengths, and their channel numbers and it will be possible to give you a much better (though not perfect without spectrum analysis) channel number to use.

By non-overlapping I meant that it's well known that the 2.4GHz band is about 60MHz wide. That permits 3 20MHz WiFi signals (11b/g and some forms of 11n). Those three are channels 1, 6 and 11.

Ideally, the WiFi alliance would have defined only channels 1,2,3 those being the three that don't overlap. But they didn't - way back, because the "occupied bandwidth" of signals like 802.11 wasn't yet agreed to - vs. now where 20MHz is the norm (excluding the hoggy 40Mhz mode of 11n).

But yes, when people choose other than 1, 6 or 11 it's not as clean.

To repeat though, the number of SSIDs on whatever channel has no real importance. What does matter is how heavily a given SSID is used. So took, signal strength of nearby SSIDs doesn't matter much. The percent of channel (20MHz) utilization as we techies say. Only a very few WiFi SSIDs have more than an average of a few percent of channel utilization. The ping times from client to WiFi router tell you if at that time of day if some other SSID on a channel + or - 10MHz of yours has a high utilization. It's very statistic - depends on your neighbors use cases and habits.

More use of Netflix et al on WiFi is the back breaker. Hopefully you don't have too many neighbors making a habit of that. If so, it's likely you can choose a different 20MHz swath to use. Or go on to 5.8GHz where WiFi is inevitably headed due to overcrowding from Netflix on WiFi.

 

[Mark Uhde]

The band is actually 80MHz wide (60MHz between the center frequency of 1 and 13 yes, but remember OFDM channels are 20MHz wide, so they extend 10MHz either side of the center frequency), and the ideal non-overlapping channels are 1, 5, 9, and 13.

There's three problems with this...

1. 802.11 legacy/802.11b DSSS channels are recommended to be spaced at least 22MHz apart, as they have a 20MHz carrier (allowing 2MHz of guard space between users). Since this isn't possible as designed (except channel 14), 1, 6, and 11 are used as the non-overlapping channels. They're 25MHz apart, resulting in a 3MHz waste of spectrum. OFDM channels have a 16.25MHz carrier and the total space recommended is 20MHz. Channel 14 is 22MHz above 11 and is DSSS only, only ever approved in Japan, and essentially useless today even there given how slow DSSS is. But the channel plan is still designed around it.

2. Based on this, much equipment ships defaulted to 6 or 11 and many people know "1, 6, 11." It's NOT good to have 5 and 6 or 11 and 9/13 used in the same space. It's much better to just stick to 1,6,11 than let this happen.

3. As far as I'm aware, no 802.11 equipment supports channels 12 and 13 when set to a US regulatory domain, or has been approved for operation on 12 and 13 in the United States. 12 and 13 are not strictly illegal, per se, but the band directly above Wi-Fi is restricted in the United States so allowing a 10MHz guard space has been the de facto regulatory requirement. Even if you decide to break the law and "risk it" on 13, you have to deal with neighbors on 11 and incompatible client cards. Why bother?

The ONLY channels should be 1,5,9,13. But that's not reality. You have to look at the radio spectrum in your area to make sure the channel you want to use has no strong adjacent channel users.

 

[stevech]

At the risk of boring the average WiFi user to tears...

The band is actually 80MHz wide (60MHz between the center frequency of 1 and 13 yes, but remember OFDM channels are 20MHz wide, so they extend 10MHz either side of the center frequency), and the ideal non-overlapping channels are 1, 5, 9, and 13.

In North America, the FCC regulations permit operations only on the channel frequencies for WiFi ch. 1-11; ch 12 and 13 are "legal" only in some other countries. (Channel numbers are WiFi alliance/IEEE defined, not FCC defined).

802.11b/g/n (excluding 11n 40MHz mode) all occupy 20MHz.

There's three problems with this...

1. 802.11 legacy/802.11b DSSS channels are recommended to be spaced at least 22MHz apart, as they have a 20MHz carrier (allowing 2MHz of guard space between users). Since this isn't possible as designed (except channel 14), 1, 6, and 11 are used as the non-overlapping channels. They're 25MHz apart, resulting in a 3MHz waste of spectrum. OFDM channels have a 16.25MHz carrier and the total space recommended is 20MHz. Channel 14 is 22MHz above 11 and is DSSS only, only ever approved in Japan, and essentially useless today even there given how slow DSSS is. But the channel plan is still designed around it.

2. Based on this, much equipment ships defaulted to 6 or 11 and many people know "1, 6, 11." It's NOT good to have 5 and 6 or 11 and 9/13 used in the same space. It's much better to just stick to 1,6,11 than let this happen.

3. As far as I'm aware, no 802.11 equipment supports channels 12 and 13 when set to a US regulatory domain, or has been approved for operation on 12 and 13 in the United States. 12 and 13 are not strictly illegal, per se, but the band directly above Wi-Fi is restricted in the United States so allowing a 10MHz guard space has been the de facto regulatory requirement. Even if you decide to break the law and "risk it" on 13, you have to deal with neighbors on 11 and incompatible client cards. Why bother?

The ONLY channels should be 1,5,9,13. But that's not reality. You have to look at the radio spectrum in your area to make sure the channel you want to use has no strong adjacent channel users.

I don't agree with the last paragraph. Signal strength and the number of SSIDs detected don't mean much. What matters to real users is how "busy" the frequencies around the intended channel are. The channel utilization.
Users should not be concerned if there are many strong signal SSIDs detected.

Consumer WiFi doesn't enable users to know what frequency ranges (channels) are too busy on average.

In lay terms, an analogy: If you had an old party-line telephone, you're all sharing the same "channel". Only if others on the line are too chatty does it affect you.

 

[Mark Uhde]

In North America, the FCC regulations permit operations only on the channel frequencies for WiFi ch. 1-11; ch 12 and 13 are "legal" only in some other countries. (Channel numbers are WiFi alliance/IEEE defined, not FCC defined).

802.11b/g/n (excluding 11n 40MHz mode) all occupy 20MHz.

Neither one is technically true. 12 and 13 are absolutely legal in the US, the problem is that emissions requirements directly above channel 13 are extremely strict. Thus, the defacto standard was just to seek approval for 1-11 in the US. But 12 and 13 could, potentially, be used at reduced power with better filtering in the radio chain.

I don't agree with the last paragraph. Signal strength and the number of SSIDs detected don't mean much. What matters to real users is how "busy" the frequencies around the intended channel are. The channel utilization.
Users should not be concerned if there are many strong signal SSIDs detected.

Consumer WiFi doesn't enable users to know what frequency ranges (channels) are too busy on average.

I don't agree with it either. I wrote it because, without the right tools, it enables you to make a best guess. The more and stronger the signals are the more likely there's heavy utilization of those channels in your airspace. Obviously, there's no absolute truth to that!

 

[Guz]

The only thing I know is that for some reason if I setup my router to use all 13 channels here in the US, several of my devices have a hard time connecting to it or have drop issues. Even if I set my router to use channel 1, 6, or 11. God forbid if I let the router chooses channels 12, or 13 on it's own, then it's even worse.

I can't explain it, but that's what I've observed.

 

[stevech]

Neither one is technically true. 12 and 13 are absolutely legal in the US, the problem is that emissions requirements directly above channel 13 are extremely strict. Thus, the defacto standard was just to seek approval for 1-11 in the US. But 12 and 13 could, potentially, be used at reduced power with better filtering in the radio chain.

This has gone anal. The US/Canada regulations require such extreme adjacent channel interference avoidance and low power, that filters to achieve that are (a) not viable in an engineering sense due to the group delay and (b) would be way to costly. So a claim that ch 12, 13 could be used for more than 50 ft. transmissions is vacuous.

The more and stronger the signals are the more likely there's heavy utilization of those channels in your airspace. Obviously, there's no absolute truth to that!

Your'e right! The truth is that signal strength is no indication of how "busy" the channel's occupied spectrum is. The words, above "stronger the signals are the more likely..." just isn't true. No more so that a car's horn loudness is any indication of how busy the freeway is.


Let's not debate this further. This minutia isn't relevant to consumers.

 

[Mark Uhde]

Dude, dunno why you want an argument, I'm just trying to be helpful. And I never said 12/13 could be useful beyond 50ft. Just that, actually, it would be kind of nice to see devices that got 12/13 approved for the US at very low power. Think a 1,5,9,13 channel plan for hotel rooms like this (I can't represent this in 3D space obviously but this is one side of a hallway, and the other side flipped on the diagonal basically):

|01|09|01|09|01|
|05|13|05|13|05|
|01|09|01|09|01|

So there's actually a huge potential application of very low power channel 13 right there, sadly, for one reason or another equipment vendors never jumped on the possibility, and it would've had to happen in the early days.

And no, the strength of a signal has nothing to do with it's utilization, but SNR comes into play here. Let's say you have 3 AP's on channel 1, and the strongest is -60dBm (neighboring apartment), whereas you have 7 AP's on channel 6 and the strongest is -90dBm. While you can't KNOW the strength of each user on 6, it's a REASONABLE BET that channel 6 is more likely to be able to get you a SNR adequate for good speeds. Of course, that's not a bet I make when doing installs, but the average consumer has to work with the tools available to them. Thus signal strength is an adequate tool to make a reasonable bet on what channel is most likely to be cleanest for them, nothing more though.