Forum

If you guys/gals haven't read this whitepaper yet, please do. I would love to get some feedback on this whitepaper.

It's called "Protection Ripple in ERP 802.11 WLANs", and it's posted on our Learning Center. You can click on the "802.11 Standars (a/b/g)" to go directly to it.

Thanks!

Devinator

Sorry not to have gotten around to reading your Protection Ripple whitepaper before now, Devin. It's not long, but it's packed with a lot of detail. And I had to get the CWNA out of the way first. :)

I really like its presentation and explaination of why an 802.11b device will degrade the performance of an 802.11g network. And how 802.11b and g stations can play dueling ripples

Your "ripple" description reminds me of a physics film I saw in school maybe 30 years ago. It demonstrated atomic chain reactions using a room filled with loaded mouse traps, each trap with a ping pong ball on it. You can imaging the sight (and sound) when one additional ping pong ball was thrown into the room. "Ripple" is a nice way of saying "unstable, oscillating chain reaction."

The one thing I'd like more information on is how the distance between APs can cause non-overlapping channels to overlap. Most 802.11 literature (including the CWNA Study Guide on pp. 77-78 ) says that channels 1, 6, and 11 have a 3MHz gap between them. Elsewhere I've read that, in fact, 1, 6, and 11 do have slight overlap. Now you say that under the right circumstances, channels 1 and 11 can overlap. Yikes! Tell me more!

at close range, at high output power, channels 1, 6, & 11 (or even 1 & 11) can overlap. It all depends on the "perceived" signal strength. Consider this analogy:

Three people are standing about 100 yards apart in a triangle formation. They're yelling at the top of their lungs. A person standing in the middle of the triangle would have no problem distinguishing what any of the three were saying. None of the three yelling people would really be affecting the others.


Suppose you bring them all to a range of 2 yards apart in that same formation, then put that same person in the middle. Now each of the three yelling people would be severely interfering with each other from the perspective of the person in the middle. What could be done to fix this problem now that the yelling people are close?

1. tell them not to yell, but to instead just talk
2. tell them to back up again

:-)

Devinator

Nice analogies. It makes our learning simpler. You got a gift there that is refreshing. I need that kind of understanding as I study for the CWSP, its seems that my brain is so saturated with EAPs and PEAPs I seep with overLEAPING RF channelization.

Your explanations are easy , simple with common everyday terms much like biblical parables. I bet you also teach Sunday School.

compughter

Devinator Escribió:

Three people are standing about 100 yards apart in a triangle formation. They're yelling at the top of their lungs. A person standing in the middle of the triangle would have no problem distinguishing what any of the three were saying. None of the three yelling people would really be affecting the others.

I understand this if the three people are three APs all on the same channel. But three APs each on a seperate, non-overlapping channel should not interfere with each other regardless of their physical proximity to each other. They are simply not able to "hear" each other regardless of how close they are to each other.

You seem to be implying that the 22MHz width of a DSSS channel varies as the distance between the APs changes. I don't understand this.

You're right that if you take that analogy that far, that it would be like they're on the same channel...but humans don't speak on different channels so it's hard to draw a straight analogy. Consider the first graphic on page 4 of my Ripple Whitepaper found at:
http://www.cwnp.com/learning_center/search.php?subcat=100&term=802.11-Standards-(a/b/g)

You have 3 APs in a room, one on ch1, one on ch6, one on ch11. These signals are not narrowband like a radio station, but rather take up a wide set of frequencies (spread spectrum (aka. wideband)). If the noise floor is moved down (or effectively moved down relative to the emitted signal strength) like it is when you're very close to a power source, then you can see why each of the APs could "hear" each other. If they were moved farther away from each other, the signal would shrink closer to the noise floor (aka. SNR decrease) yielding less "overlap". The same would be true if the output power of the transmitted signal were to be decreased or an antenna with less gain were used. It's the SNR that we're concerned with here.[/url]

I think essentially what you are saying is that the guard bands were not sized properly when designing the standard. When the ap's are close together since the signals are stronger they are able to bleed through just enough to be heard by an ap on a different channel.

This leads me to ask who or what group is in charge of conducting testing on the interoperability of new and existing standards?

Just our of curiousity did you perform the test with multiple vendors ap's?

That was my post above.

Well, as far as "interoperability", the only organization doing that is the Wi-Fi Alliance. When it comes to making rules about what you can/cannot do with WiFi devices, that's up to the FCC. When it comes to what the "industry" is agreeing to do as a whole, that's the IEEE Standards. You have the right concept in mind. I did the testing with same and different vendors - same results. Apparently more than one company is doing things this way. :-) Yuck. Ripple can be used as a DoS attack. hideous.

Devinator Escribió:

Ripple can be used as a DoS attack. hideous.

Whoa. I don't remember that explicitly mentioned in the whitepaper. Maybe you should add a sidebar on that fact and give a more detailed "how to" on implementing a way to reduce or prevent ripple.

I've seen "NO BLUETOOTH" signs in areas with 802.11 WLANS. Now I'm expecting to see "NO 802.11b" signs in areas with 802.11g networks.

In fact, is there a whitepaper on 802.11 network-specific attacks? Jamming using Bluetooth (FHSS) devices tops the list. The ripple effect is an excellent example of another. What others are there?