Forum

Hi rhinsch,

There's really no such thing as a "pure 40 MHz" mode. There are three bandwidth modes in 802.11n:

20/40 MHz, 20 MHz, PCO

Hopefully this article will help with this:

http://www.cwnp.com/community/articles/802.11n_20_40_mhz_bss_mode_operation.html

There are 4 operational modes in 802.11n:

Modes 0-3, each being used for a specific purpose. The system automagically switches back and forth between them as necessary.

There are three PPDU formats in 802.11n:

http://www.cwnp.com/community/articles/802.11_ppdu_formats.html

Now, there is definitely such a thing as a greenfield deployment, but in the real-world, it basically doesn't exist. You *could* manufacturer this scenario by renting an abandoned building 2 miles out into a wooded area where no other Wi-Fi exists and then using only 40 MHz 802.11n capable devices. 20 Mhz transmissions will almost always be present due to nearby legacy Wi-Fi or legacy Wi-Fi that's part of your network. Also, beacons are sent out using legacy frame formats at legacy PHY speeds (1 Mbps DSSS for example).

So the answer is yes, your a/g WIPS will see that 802.11n network (whether APs or STAs), but won't understand that it's an 802.11n unit. It will only see frame types (and speeds) that it understands and will assume that it is either 802.11a or 802.11g or both.

I haven't read Joshua's preso (but I will today), but if he says it, you can basically bank on it. ;)

I assume that when an AP is manually set for Greenfield mode, that it uses the Greenfield PPDU format for everything including beacons (I need to look that up to be sure). Additionally, STBC (space time block coding) can be used for beacons and other management and broadcast frames, which means that a/g sensors wouldn't understand what they're looking at. STBC has its own Basic MCS rates, so if you're sending at at a unique data rate while using a unique frame format in a manner that splits the data transmissions into multiple pieces, it's going to completely escape detection by a/g sensors. That's alot of IFs, but under all of these circumstances, it would work.

Devinator