3com WAP - public/secure and 'N' vs B/G

here's some info I found - but it discusses the client side, not the AP side -

If they are running an 802.11n network they don't have to allow 11b or

11g to connect, the draft 11n only says that the equipment must be capable of connecting to legacy devices so the user can have the capability to run a mixed network. Look at the set of 11n mode options that are available on this emulator

If the access point is setup for 802.11n - and only 802.11n - would I still see the SSID beacon on my b/g card via NetStumbler, Hopper, Windows ?

Is the beacon packet seen by all technologies at the lowest common "speed & structure", and only goes into 802.11n mode after the actual connection ?

Would I see the SSID, but just not be able to establish a connection ?

Without actually trying it I am not sure exactly what you would get. I have been doing a bit of reading about N recently and am not sure how some of these "legacy" connections are supposed to work. "During the definition of the 802.11g standard, it was realized that, since legacy 802.11b devices would not be able to decode the newer

802.11g frames, there ought to be a separate mechanism to help legacy devices set their NAV correctly and therefore to reduce the percentage of collisions on air. The 11g standard made use of existing ?protection mechanisms? ? RTS and CTS ? to help legacy stations set their NAV. A similar situation arose during the definition of the 802.11n standard. Legacy 802.11a/b/g devices would not be able to decode the 802.11n headers ? and therefore a protection mechanism becomes necessary. One of these is the transmission of legacy preamble and header that enable the 802.11a/g/ device to detect the 802.11n packet and to decode the information in its signal field, from which the correct packet duration can be determined."

My reading of that is you should see something.

There was also this "Backward compatibility with legacy devices also may be enabled by forcing devices that are compliant with a newer version of the standard to transmit special frames using modes or data rates that are employed by legacy devices. For example, the newer devices may transmit Clear to Send/Ready to Send exchange frames or Clear to Send to self frames as may be employed in standard 11g. These special frames contain information that sets the network allocation vector of legacy devices such that these devices know when the wireless channel is in use by newer stations."

I have heard mixed reports about even using an 11n card with netstumbler with some people not getting it to work and one person who got it work only having it report speed at 54Mbs.

I recently borrowed a wireless router that accepts cell phone 3G wireless cards (this one was Sprint) and rebroadcasts the signal as standard b/g/n (I don't recall the router brand)... I wanted to see if this setup had enough bandwidth to allow three b/g equipped laptops to surf the net without obvious slowdowns (it did). The default was set to broadcast in b/g/n modes, so at one point I selected "n only" and after the router restarted all three laptops detected the network but could not connect. I then connected to the router with a network cable to change the setting back...

Which N mode? Beam forming or spatial diversity? In beam forming (Ruckus Wireless), it's basically the same as 802.11g and you will see beacons. In spatial diversity (Airgo), I think (not sure) that the spec demands beacons. However, I'm fairly sure that the beacons are at 6Mbits/sec as in 802.11g. If you enable 802.11b compatibility, beam forming still works and it will belch beacons at 1Mbits/sec. However, with spatial diversity, 802.11b and MIMO speeds are mutually exclusive and I really doubt that you'll see 1Mbit/sec 802.11b beacons. No clue what happens with the new chips that do both beam forming and spatial diversity. I gotta read the specs to be sure, but I think that's the way it works.

What you're basically asking is how does 802.11n operate in "greenfield" mode, where there are no constraints imposed by prior technology and legacy junk compatibility. Greenfield mode is very unlikely to be seen in a public library. My guess(tm) is what you're seeing is your wireless client software annoucing the highest level of acronym compliance. Just because you can't connect doesn't mean the lower standards are not available or functional. It would be nice if Netstumbler and others annouced all the various supported 802.11 acronyms, but they don't.

Incidentally, se have a somewhat similar derrangement at the local hospital. You can connect, but can't do anything without a WPA password *AND* and a VPN client. Everyone complains that they can't connect, until they finally read the splash page. Anyway, don't assume that it's 802.11n "only" just because you can't connect.

I don't know if this really helps, but the section on backwards compatibility:

gives a few clues on how much of 802.11g technology (beacons) is compatible with 802.11n. I'm to busy/lazy right now to find a better reference.

Gone to waste the whole weekend:

Having got my XP laptop put back together and eventually finding someone who has an 11N only network I ran a few programs. I have used only b/g wireless cards. Kismet in Ubuntu reports an 802.11n network. XPSP2 laptop running the following:-

1.WZC reports the SSID. 2.Wifihopper reports SSID and an 11g network with the selected "40MHz" channel. The graph function works as well. 3.Inssider reports SSID but the graph function doesn't work. 4.Netstumbler was erratic to say the least. It wouldn't reliably report the SSID although it did report the correct MAC address. The graph would not work, you got a single line at the start and that was that. I later found that if the 11n "standard channel", assumption on my part, coincided with the channel that one of the nearby AP's was using then the graph stopped however if that AP was then switched off the graph would resume, it was pure luck that that AP was switched off.

Vista SP1:-

1.Vista "connect to a network" shows the SSID. 2.Vistumbler 8.1 showed SSID and that it was an 11n network and the "40MHz" channel. 3.Inssider showed SSID and the "40MHz" channel. The graph function worked in Vista. 4.Xirrus WiFi Inspector showed SSID and that it was an 11n network. It also showed 2 channels,"wide" and "standard" I assume, which is why I went back and checked Netstumbler again and found the graph was now working and one of the AP's was no longer transmitting. 5.None of the versions of wifihopper work correctly with the atheros b/g card in my Vista laptop. The latest version 1.2 build 2008-110600 reports all networks as type 11a, so it works after a fashion but didn't try it as there may be other bugs. 5.Inssider and Vistumbler are based on "netsh" commands so I didn't try that either.

None showed speeds greater than 54Mbs.

I have assumed that the 2 channels picked up by the Xirrus wifi inspector is correct and not a fault with the 11n AP or the Xirrus tool. The second channel I have assumed to be the "standard channel" and this does cause a problem for netstumbler if another AP is using it as well. I would be interested to know if anyone else has used this tool and noticed whether they have picked up 2 channels with an 11n network.

since you could see the SSID - what happens when you try to connect to the 802.11n-only network ?

and how did Hopper display the 802.11n with the 40mhz channel ? that would be my giveaway for a N-only network.

It's not my network and it was encrypted so I couldn't connect. Joe M in his post reported that while he could see the network he could not connect when he tried. I know a person who works at the company and he assured me it was 11n only which is why I sat in the car park and ran a few tests.

You must have missed it.