Channel 13, F*#k Yeah!
Wi-Fi is a technology used around the world and we, as Americans, respect all cultures and creeds. That said, occasionally we like to celebrate ourselves (NSFW, as if the title didn't make it obvious):
Unfortunately, America has been behind the times in Wi-Fi. No longer! We now have motherf*#king 2.4 GHz channels 12 and 13.
This blog has a rigid rule to avoid politics, so I'll phrase this as apolitically as I possibly can: from January 20, 2009 to January 19, 2016, America's wireless authority -- the Federal Communications Commission (FCC) -- repeatedly authored rules that damaged Wi-Fi. In 2010 the FCC banned Wi-Fi on channels 120 through 128, in 2014 the FCC messed up channels 52 through 144 by requiring additional dynamic frequency selection (DFS) scanning and in 2016 the FCC declined a proposal by satellite phone provider Globalstar to allow 2.4 GHz channels 12 and 13 to be used for "terrestrial, low power services" (TLPS), which just so happen to have the exact same rules and protocols as 2.4 GHz Wi-Fi.
In fairness to our beloved/reviled former President's FCC, channels 120 through 128 were re-instated in 2014 and the DFS scanning rules don't effect most well-designed enterprise Wi-Fi networks because a well-designed Wi-Fi network doesn't use DFS channels (future blog topic alert!).
Today America has a President whose FCC is much friendlier to Wi-Fi. Less than a month after the man who likes Real News better than Fake News took office, the FCC approved Globalstar's petition, thus opening up channels 12 and 13 for 802.11n (2.4 GHz) Wi-Fi networks.
As with many things in today's USA, the expansion into channels 12 and 13 wasn't a simple issue. The Globalstar proposal to Make 2.4 GHz Wi-Fi Great Again was controversial because Globalstar continues to own those channels. Wi-Fi AP and device makers would have to pay Globalstar a small fee for the right to use those channels. (Settle down, do-gooders. Lots of private companies get paid by Wi-Fi AP and device makers already. It's an essential part of the 802.11 Standard sausage-making process.) There has been some question as to whether AP and device makers (especially device makers, because they drive the industry) would choose to use Globalstar's semi-unlicensed channels.
During some down time during my travels today (which is my carny way of saying that I'm stuck in an airport), I decided to check to see if 2.4 GHz channels 12 and 13 were being adopted by Wi-Fi devices.
I set my MacBook to capture on Channel 13 (2472 MHz). Lo and behold, I saw Wi-Fi traffic:
While the presence of Beacons & Probes and assorted management & data frames is compelling evidence that Channel 13 is being used, it is not proof. The channel frames are captured on is not always the channel that frames are transmitted on. In other words, the Wi-Fi traffic in this airport may have been transmitted on Channel 11, but captured on Channel 13.
How, then, does one determine whether Wi-Fi traffic captured on Channel 13 was transmitted on Channel 13? By looking at the RSSI of Probe Request frames. When Probe Requests are captured on a different channel than they are transmitted on, the RSSI of the captured Probes is artificially low.
So, I captured on Channel 13 and isolated the Probe Request frames from one device:
Then I captured on Channel 11 and isolated the Probe Request frames from the same device:
On both Channel 13 and Channel 11, the highest RSSI of captured Probe Requests was slightly below -50 dBm. Since the RSSI of frames captured on Channel 13 was approximately the same as frames captured on Channel 11, that told me that the device was almost certainly Probing on Channel 11 and, importantly, Channel 13.
While it is true that one Wi-Fi device in one airport using Channel 13 is no guarantee of coming widespread support for TLPS, it is a sign that Channels 12 and 13 could be coming to America to save the motherf*#kin' day for 2.4 GHz Wi-Fi.
If you like my blog, you can support it by shopping through my Amazon link. Thank you.
Twitter: @Ben_SniffWiFi
Twitter: @Ben_SniffWiFi
ben at sniffwifi dot com
F*#K YEAH!!! Just 1 more until we get an extra non-adjacent channel!!!
ReplyDeleteSomeone pointed me to the radio-tap header using omnipeek once - identified the channel for me without a doubt. It was actually because I had the phone I was testing next to my laptop while I was sniffing frames on another channel, I asked a dev why I was seeing it's frames on my sniffer channel while the phone was supposed to be on another. He showed me the radio-tap header and taught me about the whole picking up frames from other channels thing.
Sorry we missed you at #WiFiTrek. Hope to see you next time!
Channel 13 gives us an extra non-interfering channel. 1 - 5 - 9 - 13
Delete"non-interfering" could you expand on this? I thought 1-5-9 would all overlap with one another, causing adjacent channel interference.
DeleteThe technical details are:
Delete1. Channels' frequency ranges and, therefore, whether they overlap are defined by center frequencies and channel widths.
2. Channel 1 has a center frequency of 2.412 GHz.
3. Channel 5 has a center frequency of 2.432 GHz.
4. Channel 9 has a center frequency of 2.452 GHz.
5. Channel 13 has a center frequency of 2.472 GHz.
6. The 802.11b PHY uses DSSS and a 22 MHz-wide channel, or 0.011 GHz on each side of the channel's center frequency. So, for 802.11b (ancient, I know!) the frequency ranges are 1: 2.401 - 2.423 GHz, 5: 2.421 - 2.443 GHz, 9: 2.441 - 2.463 GHz, and 13: 2.461 - 2.483 GHz. These do overlap slightly (2 MHz of 22 MHz for 1 and 13, 4 MHz [2 MHz on each end] for 5 and 9).
7. The 802.11g and later PHY use OFDM and 20 MHz-wide channel (or 40 MHz in the case of 802.11n on ISM), or 0.010 GHz on each side of the channel's center frequency. In this case, the frequency ranges are 1: 2.402 - 2.422 GHz, 5: 2.422 - 2.442 GHz, 9: 2.442 - 2.462 GHz, and 13: 2.462 - 2.482 GHz. While these do not technically overlap, this is only in the theoretical case of a perfect bandpass filter that does not permit any "spill" outside of the nominal frequency range, there is no "guard" band separating the channels, and it does not take into account the lower amplitude but wider side lobes of the main RF signal.