"interference robustness"?

[Doug Lassiter]

The 3.1.1 driver for Airport base stations, released last year, offers a check box for "interference robustness" in Airport Admin Utility. This is advertised as a way of getting around GHz interference from phones and microwave ovens, and near as I can tell from my own experience and that of others is that yes, it works. What is it? More pointedly, what do I lose if I check that box, and why does Apple see the need to give the Airport user an option to choose to have non-robust wifi in the first place? For example, does this option cap the bandwidth? If so, it'd be nice to know.

On the basis of fifteen minutes of googling, it sounds like this is a recurring question, and that Apple has offered little or no info. The most frequent explanation is that "interference robustness" makes your system more robust to interference, which the posters seem to consider an insightful piece of information.

Some thoughtful detective work is reported at

http://www.neato.org/~page/virtmem/archives/000070.html

though the suggestion that reduction of receiver sensitivity makes performance better is slightly non-intuitive.

Ideas? Info? Experiences? My apologies if the forum has discussed this. I wasn't able to find it.

[G5from128k]

I'm surprised nobody has taken the 15 minutes need to test the effect on bandwidth. Copying a few gigabytes across the network with it on or off while watching Network Activity on the Activity Monitor could provide some useful insight. Interference Robustness definitely helps cope with interference and a simple test might help determine if it has a hidden downside.

[Doug Lassiter]

Yes, I did this, but with results that were pretty inconsistent. I assumed that was because I did this on a network with a fair number of machines on a single AE router, and I wasn't able to tell everyone to quit using the network. The differences that I saw weren't large, though, and that's why I think there is more to it than just bandwidth. Clearly one can do an experiment here, but one needs a simple enough setup to make it meaningful. I should have an opportunity to do this right in a little while, but was just wondering if there was some complete explanation available. I do see indications that "Interference Robustness" is indeed a microwave communication standard, and not just Apple handwaving, though.

[Doug Lassiter]

I posted this question to alt.internet.wireless, and got a useful response from Jeff Liebermann. there, who seems to know something about microwave communication. Thanks, Jeff!

Although Apple doesn't seem to want to tell anyone about interference robustness, Orinoco (Lucent), who contributed strongly to Airport design (I think Airport at one time used their PMCIA card) is willing to help us. See

http://www.orinocowireless.com/support/techbulletins/TB-035.pdf

The answer is that Orinoco provided a settable option in their PMCIA to give robustness against microwave ovens. We believe that this option is the one that Apple calls "interference robustness". The key to this option is that microwave ovens have a strong periodic component to their radiated power. For 60Hz ovens, at least,, they spend 8.3ms radiating strongly, and 8.3ms being quiet, out of a 16.6ms period. Now, if a PMCIA sees interference, it'll slow down the transmission more and more. At some point, the packet size takes longer than 8.3ms to get out, and such long packets will get whacked every single time (at least by a 60Hz oven) - not just occasionally. What the option does, I think, is to prevent the PMCIA from ever backing off to packet sizes that are longer than this. Clever. Basically keep the PMCIA from shooting itself in the foot as it valiantly tries to keep up.

The bottom line is that the option shouldn't make much of a difference unless ACK failures are common. That's when packet management becomes an issue. For truly random noise sources, in fact, keeping this option off may be the best noise mitigation strategy. So my guess is that turning it on to kill 2.4GHz telephone noise or rogue wifi router noise might actually be disadvantageous. But do what works.

[bangalla]

All I can add to this is that I have found interference robustness to have a profoundly positive effect on my TiBook. My Airport is located at the other end of the house near an exterior wall and the TiBook would constantly drop out. I was continually harassed by the sound of iChat logging in and out.

Since activating interference robustness I get a steady two or three bars, where before it would vacilate between two and nothing at all. YMMV, but I'm overjoyed that I now won't have to get another airport for this end of the house, or add an antenna to my existing one.

Dave

[marvinmayhem]

I live in an apartment in a high population density area. My ibook is usually about 10-15 feet from my airport express base station, and I generally get 5 bars signal strength. When I first got the airport, I was having some problems with the airport dropping out completely. I could still see other wireless networks, but not my own. The only way to get it back was to powercycle the base station. I called Apple Support, they suggested turning on interference robustness, and the base station worked fine for a few months. Lately, though, it seems to drop out every night. In the morning, I need to powercycle it to get it working again. I've tried various combinations of interference robustness on and off, changing the channel, and the multicast rate. I've reset the WEP password, thinking someone might be hacking into it at night, and recently, using the Admin Utility I restricted access to the base station to my MAC address. Any ideas what might be causing the problem, or ideas how to fix it?
Strangely enough, there seems to be a similar problem with cellphones in my apartment- they get full signal strength, but abruptly drop connections, usually within a few minutes of making a call. I don't use a cordless phone or microwave. Thanks.

[yellow]

Any ideas what might be causing the problem, or ideas how to fix it?

Any of your neighbors could be using a cordless phone or a wifi accesspoint with the same channel. That would cause all sorts of interference.

[unother]

I've reset the WEP password, thinking someone might be hacking into it at night,
Ack! What are you doing, man? Why even use WEP? Use WPA-PSK--which aside from being the recommended method, means no crazy hex passcodes...
Strangely enough, there seems to be a similar problem with cellphones in my apartment- they get full signal strength, but abruptly drop connections, usually within a few minutes of making a call. I don't use a cordless phone or microwave. Thanks.
Sounds to me like you have a high-level interference from something in your neighborhood. If you're near a power generation facility...

[voldenuit]

Marvin, it could be helpful to use stumbler software to find out what channels the strongest neighbour-WLANs use and to keep clear of them. And WEP is indeed obsolete.