[Simon]

So let's say you live in a city and there are many wireless networks around. You want to chose a channel for your network that is previously not being used. Now of course you can use a tool like AirMoose or iStumbler to see which channels are being used by other networks, what the signal strengths and noise levels are, etc.

But what about networks that hide the SSID? Thos won't even show up in these apps and yet the channels they're using won;t be suitable. And what about other noise sources on those channels? Is there any tool that can scan these channels and report on amount of "traffic" or noise on that channel regardless of the type of network running in that frequency range?

[ghporter]

It looks like AirGrab will identify those "closed" networks for you. I haven't tried it, so if you do, post how it works for you.

Would one of those hand-held WiFi sniffers work for this? I haven't tried any of them either...

[Simon]

Thanks for that link.

Unfortunately the app won't even launch. And this is on a brand new MBP running 10.5.2.

[Big Mac]

I use Network Stumbler on my Dell.

[dimmer]

The concept of multiple wireless networks having to co-exist was a known part of the specification process, and the "Automatic" selection for both the base station and the clients should be optimal (and of course adjust to any new networks that appear, or old ones which go away). While it's possible to identify a specific alternate channel, that turns into a crapshoot when other folks in your area do the same. Equipment that requires specific channel selection is way below standard.

"Blocked" or "Hidden" SSID's don't block the traffic on whichever channel they use: they just don't give a public ID of the network name. Again, an automatic selection will work best around these, and is the closest to "zero configuration" possible.

[ghporter]

Originally Posted by dimmer 

The concept of multiple wireless networks having to co-exist was a known part of the specification process, and the "Automatic" selection for both the base station and the clients should be optimal (and of course adjust to any new networks that appear, or old ones which go away). While it's possible to identify a specific alternate channel, that turns into a crapshoot when other folks in your area do the same. Equipment that requires specific channel selection is way below standard.

"Blocked" or "Hidden" SSID's don't block the traffic on whichever channel they use: they just don't give a public ID of the network name. Again, an automatic selection will work best around these, and is the closest to "zero configuration" possible.

In practice, the "Automatic" setting is often way off optimal-I don't know what algorithm is used, but it seems to wait for a new network to be there for a while before it decides to change, and from what I gather, ANY traffic on a channel means that it's marked as in use, whether the signal is strong or weak. Assuming that there was a truly optimized algorithm to determine what was going on and appropriately respond, then this setting would be fine. But saying that anything that lacks this Apple-specific feature is "way below standard" is disingenuous.

While hidden or "closed" networks don't "block" traffic, they still occupy bandwidth, which CAN interfere with one's own network traffic-rather badly. It's VERY important to know what channels are being used to ensure that you choose the channel that's least likely to be interfered with.

[mduell]

Yea, NetStumbler is great for finding channels for networks that don't broadcast their SSID... but I haven't seen anything as capable for OS X.

[dimmer]

There is nothing Apple specific about the protocol: it's all part of the IEEE standard. And yes, it will take into account "hidden" networks: believe it or not, network engineers do, sometimes, think it out.

While hidden or "closed" networks don't "block" traffic, they still occupy bandwidth, which CAN interfere with one's own network traffic-rather badly. It's VERY important to know what channels are being used to ensure that you choose the channel that's least likely to be interfered with.

Actually, it only matters if other users are so anal they use specific channels, rather than letting the technology handle it for them. Which includes "hidden" and "blocked" networks. The optimal performance test takes place with every packet sent, so it will always perform as well as a fixed channel (usually much better).

[ghporter]

Originally Posted by dimmer 

There is nothing Apple specific about the protocol: it's all part of the IEEE standard. And yes, it will take into account "hidden" networks: believe it or not, network engineers do, sometimes, think it out.

While hidden or "closed" networks don't "block" traffic, they still occupy bandwidth, which CAN interfere with one's own network traffic-rather badly. It's VERY important to know what channels are being used to ensure that you choose the channel that's least likely to be interfered with.

Actually, it only matters if other users are so anal they use specific channels, rather than letting the technology handle it for them. Which includes "hidden" and "blocked" networks. The optimal performance test takes place with every packet sent, so it will always perform as well as a fixed channel (usually much better).

Can you fill me in on whom other than Apple offers this feature? I haven't seen it on any non-Apple equipment.

[dimmer]

Pretty much anything 802.11wireless compliant does or at least can. Many of them ship on a set channel 1 (Linksys, I'm looking at you) but selecting 0 rather than 1 should make them behave (check the manual -- it should let you know).