Slashdot Mirror
Wireless-Friendly Microwaves
Makarand writes "According to this article on ABC News, scientists at the University of Michigan in Ann Arbor have stumbled upon a simple and elegant solution to keep your kitchen microwave from
becoming a noisy nuisance to your home Wi-Fi network.
They found that they could focus the microwaves into a single frequency and
reduce noisy
microwave emissions by placing ordinary magnets in specific patterns along the magnetron .
New techniques to reduce microwave interference will be needed when
Wi-Fi enabled entertainment systems will allow digital audio and video to be transmitted
to different rooms of a house wirelessly. Packet drops in such a sytem would degrade the video and audio
experience."
Low-noise microwave magnetrons by azimuthally verying axial magnetic field - here
$ strings FTP.EXE | grep Copyright
@(#) Copyright (c) 1983 The Regents of the University of California.
The article stresses that those microwave interferences can be curtailed with "ordinary magnets" placed "in a specific pattern" so why isn't there a DIY guide for figuring out that pattern and slapping the magnets on the side of the oven? I know I'm probably oversimplifying, but if you know the pattern at which your oven emmits the microwaves, it can't be too hard to figure out the pattern at which you can put the magnets. Am I missing something? Or is it simply because, as they mentioned, reducing microwave interferences is a huge market and "opensourcing" the method would stop that?
I have big problems with the radio that the CIA implanted in my brain.
They found that they could focus the microwaves into a single frequency
...
Why not go all the way and make the frequency and phase of the microwave oven's magnetron adjustable, add some kind of microcontroller to drive it, and a small cpu to implement the 802.11b stack. Then, from your laptop, run this script:
WIFI_IF=eth0
DATE=`date +%s`
while [ ! $TIMEOUT ];do
DATE_PREV=$DATE
tcpdump -i $WIFI_IF -c 1
DATE=`date +%s`
let TDIFF=DATE-DATE_PREV
if [ $TDIFF -gt 5 ];then
TIMEOUT=1
fi
done
echo "Coffee is hot!"
Ah, the marvels of technology
"A door is what a dog is perpetually on the wrong side of" - Ogden Nash
More like be prepared to be booted as the microwave will cut off 802.11b. I did not have a problem with this until I bought my current microwave (it's a bit closer to the AP then the old one was). When ever the microwave is on, the signal drops to zilch and I disconnect. Makes me wonder how safe these things are! :)
Gorkman
legacy microwaves
Now there's a phrase you'd only hear on slashdot.
Remember, it takes 42 muscles to frown and only 4 to pull the trigger of a sniper rifle.
Microwaves DO interfere with WiFi. Case in point, when my wife is cooking here egg rolls in the microwave, my WiFi signal drops to zero. The microwave finishes and poof...11mb connections. If I can connect, I either get really really really SLOW connections, or I have to be ontop of the AP....literally! 2.4 GHz is what many call the garbage band...you got cordless phones, cellphones are close to it, microwaves, WiFi (both a and g), video units, intercoms, and just about anything you can think of all fighting for spectrum. 5GHz is going to be no better. I am waiting to see if the either start cleaning up 2.4 GHz which would be REALLY hard, or ramp up or down the frequency. I thing the 1.2 GHz ham band would be a good candidate for refarming. From Ham use, it's not even close to being useful for public service and even if there are radios, there's usually noone there even during rush hour. The range would be a bit better then 2.4Ghz and they could totally reserve it for WLANS of all types. As a ham, I am not usually in favor of killing a band (more in favor of addding ham bands), but almost no friends of mine work 1.2GHz and I am sure all of them would like a better WLAN connection! ;)
Gorkman
When you hit water with it, it will agitate the molecules and things get hot and cooked
You seem agitated and cooked enough without microwaves.
The difference is in power and concentration : a microwave oven is minimum 700W, concentrated on a lump of water, whereas an 802.11b is 100mW radiated in all directions. You'd need hundreds of wifi cards doing denial-of-services around a cage to even start incommodating the hamster inside.
"A door is what a dog is perpetually on the wrong side of" - Ogden Nash
There may not be anything WRONG with the microwave. This is BS. First off, actually at power leakage is probable not occuring. It's REALLY hard to keep RF from getting out. First, you HAVE to have a door to put food in and take it out. You also have to have a window (because you need to see if your food is about to turn into flames). Most leakage could come from those two areas. If I can get a signal out of a public bus at 2m, I bet that even a nicely shielded microwave could have SOME leakage that could interfere with WiFi. Both of my microwaves are less then 3 years old. Only the one that is closest to the AP (about 10ft away) causes any issues with connectivity. Also, keep iin mind, most microwaves are pushing around a kilowatt or higher PEP output (if that's how they measure it). Most WiFi AP's are 1 watt or less. Two things I may try to increase my AP's profile:
Add a Linksys Stackable Amp (called a Signal Booster, but it's just an Amp) for 99 bucks.
Get rid of the rubber ducky antennas and either build my own halfwave antenna, or by a Diamond. Not known to many WiFi guys (except the serious ones), rubber ducks that ship with pretty much any radio equipment are usually compromise antennas. The typical SWR is closer to 2 then it is to 1 and it may be higher. The lower the SWR, the more power you are radiating. Rubber ducks are included because they do work, but they are probably not the best antenna you can get. Go spend some money and buy a Diamond antenna or find a homebrew design you can make that satisfies your requirements. After designing it, make sure to use a length of low loss coax and mount it near the ceiling.
Those two things would raise the profile or your AP and maybe not eliminate, bur probably reduce the amount of interference you get since your AP is now radiating more RF. I don't reccomend building your own amp. Stick to off the shelf as you'll be sure to be within the FCC power regs. If these regs state ERP, then be careful of your antenna also. To much gain may push you over that reg and while the FCC probably won't come to your house, it's being a good spectrum user to follow those guidelines.
Don't add a directional antenna unless your trying to establish a link say from building to building, or if you have an AP with Omnis and are setting up another AP to increase your footprint in a certain direction. Omni's would work better for most situations unless your trying to establish that link or establish a lobe in a certain direction. If the general idea is to provide better overall coverage, directional antennas like a yagi are not what you want. Go with a good omni.
Gorkman
I am sitting here testing a forward error correcting file transfer protocol and in a fit of boredom I turn to Slashdot. This article is very timely as I needed a way of injecting some noise into the system.
Test environment:
IBM T-21 laptop with Orinoco gold 802.11b wireless PCCard.
SMC di-pole wireless AP (Forget the model number) which in testing has turned out to be a very good AP with range exceeding all of the standard 802.11b AP/routers we have tested.
The test file is 4MB in size and we are sending it in both directions across the wireless network with and without error correction. No suprises here, with a perfect signal the file with error correction takes slightly longer to arrive due to the increase in size. Transfer rate is about the equivalent speed to a network file copy and slightly faster than ftp on the same network.
After reading the article I moved the laptop to within two feet (as measured from microwave to the antenna of the access card) and re-ran the tests.
With the microwave off, all tests ran as normal, with the microwave on I get the following results.
Network file copy: Failed with network timeout, network not available
Our FEC file copy: completed but very slow
Our Non-FEC file copy: failed due to loss
Time to look closer. I fired up the Orinoco client tools for site monitoring which allow you to view various network conditions. With the network off the signal was typically at -72db and the noise was measured at -92db. With the microwave on the signal would range between -72db and -60db and noise would range from between -90db and -63db. With the microwave on the signal quality would range between non-existent and 'good'.
Running our tests produced the following results.
Microwave off:
-------- Transfer Summary --------
Data bytes: 45638341
Elapsed time: 91.93 seconds
Effective rate: 3971.44 Kbps
Packets lost: 11
Packets sent: 46853
Requested Rate: 10000
Actual Wire rate: 4370.70 Kbps
Average loss: 0.02%
Average RTT: 35.88 ms
Microwave on:
-------- Transfer Summary --------
Data bytes: 45638341
Elapsed time: 390.71 seconds
Effective rate: 934.47 Kbps
Packets lost: 3225
Packets sent: 50067
Requested Rate: 10000
Actual Wire rate: 1098.95 Kbps
Average loss: 6.44%
Average RTT: 85.03 ms
The two important numbers are effective wire rate and packets lost. Keep in mind that repeated attempts at shell based file copies failed completely as did a non-fec file copy using udp and tcp. This looks like a problem that really does need a solution, at least for 802.11b.
Oh, and my microwave is a two year old top of the line KitchenAid built in so it is surrounded by an additional metal frame and all of the wooden cabinets (and whatever they contain). Even with all that extra shielding it was massivly effecting the wireless throughput and presumably anything else within range, scary, I won't be standing too close to the microwave from now on when its on thats for sure.
2.4 GHz is not the resonating frequency of water. That's way way up in the GHz chain. 2.4 GHz was chosen because that band is the junk band in which unlicensed users are subject to interference as part of the spec.
Microwaves work by oscillating water molecules, which are dipole. The magnetron cycles 2.45 billions times per second, which twists the water molecules. The interior of a microwave oven is coated with a microwave-reflecting material which allows a single beam to essentially paint the three-dimensional interior.
So many people write that water resonates at 2.4 GHz. It's just not true. Here's a nice explanation of how it works.
Freelance tech journalist for the Economist, MIT Technology Review, Macworld, and others