Ofcom works out why Wi-Fi doesn't work An Ofcom-commissioned report into Wi-Fi performance concludes that it's baby-listeners and TV-senders that are mucking with the signal, not to mention the "Free Public Wi-Fi" virus, without which we'd all be connecting faster. Ofcom's remit is to ensure efficient use of radio spectrum, including the unlicensed 2.4GHz band used …
Wikipedia, natch..
"A microwave oven works by passing non-ionizing microwave radiation, usually at a frequency of 2.45 gigahertz (GHz) (a wavelength of 12.24 centimetres (4.82 in), through the food".
As I doubt if the ovens are linked to a frequency standard (mine aren't) or are crystal-controlled (I've got to whip one of mine apart 'cos of a funny sound - pictures at 11 - to discover this fascinating fact), they might stray into the WiFi band. Or just block the WiFi receiver. Probably go through the front-end like a dose of salts. That'll fuc*k the transfer rate.
Bet there are more microwave ovens in London than Bournemouth. Like, there's more yokels in B'mouth than the Smoke. Maybe.
Another fascinating nibble from Wikipedia...
"Cooking food with microwaves was discovered accidentally in the 1940s. Percy Spencer, a self-taught engineer, was building magnetrons for radar sets with the company Raytheon. He was working on an active radar set when he noticed that a peanut chocolate bar he had in his pocket started to melt."
I wonder if just the nuts in the peanut bar were the only ones affected? Again, Wikipedia "Spencer was married and had three children, James, John, and George.". Before, or after this magnificent discovery, one wonders?
Paris, 'cos she could melt my nuts.
This is hardly surprising really, the whole point of deregulated bands were to encourage lots of cheap *non-critical* radio use.
If you need any sort of quality of service then you need a well regulated system where both the technical standards and the implementation are both properly validated and enforced. That is not to say they the regulation can't be flexible to accommodate improved services, but any flexing must be appropriately controlled as well.
Anyone relying on 2.4GHz 'free' access is asking for trouble, fine for limited home use but not for important services. If there are any professional idiots out there using 2.4GHz for industrial control or similar potentially safety related projects, have you considered how easy it would be to modify a microwave oven to saturate your whole factory's RF links and do DoS attack?
What then?
While it's true that microwaves use 2.45 GHz non-ionizing radiation, that radiation is generally (or rather, should be!) confined to the interior of the oven. The metal of the case should prevent it leaking, and the reason for the metal grille in the door is that the holes are significant smaller than the wavelength of the radiation, which should prevent its escape.
Or people could start switching to 5GHz Wi-Fi... there are loads of channels up there and it's virtually unused!
As for the comment re microwaves; Look up "faraday cage", then you may realise that even though microwaves use 2.4GHz, they don't interfere with wi-fi much; it also explains why their operators don't get cooked. (hint: the 2.4GHz radiation is mostly contained within the device).
What the report doesn't make clear is whether the frame percentages are "available" frames, or "used" frames - I imagine most wifi is unused 16 hours/day, but still beaconing; hence the high percentage of beacons?
It seems that the key phrase for the 2.4GHz WiFi instances is that they operate in an unregulated part of the spectrum. We discovered long ago that once a service is deregulated (or was never regulated in the first place) you effectively get something akin to the wild-west. Just look at how bus deregulation "improved" transport services.
While I am not a fan of pointless controls, laws and restrictions there is a place for laying down some rules about what can and cannot share a scare resource, or even for standardising on methods of access (or even the sizes of things: look at the benefits of everyone using the same mains voltage, or A[0-6] sized paper, for example)
Fortunately, with WiFi there is an obvious and easy solution - switching to 5GHz 802.11tg, albeit at a higher cost and reduced range. Although that in itself helps in reducing the amount of overhead traffic.
My parents owned one of the VERY first microwaves, the Amana Radarange. It had two big circular dials, coolly backlit, that functioned as the analogue timers. No programming, no digital readout, no rotating trays.
It was built like a TANK, heavy as hell, and with a door that slammed shut tightly and hermetically.
Microwaves SUPPOSEDLY are totally sealled units, with heavy EM shielding surrounding the cavity, and doors that function as part of that shield. The Radarange had a perforated sheet of metal that was embedded in the glass of the door - today's units don't seem so well built or sealed, probabably because they cost a tenth of what the originals did. I suspect they leak a bit more EM too...
Paris, because microwaves are probably the only oven she has ever had to operate...
Completely agree about the interference - it's obvious really and why my "cheap AV sender" hit the bucket in favour of installing a few runs of co-ax round the house about 3 days after I bought it.
The stats are interesting though - they're presenting "mean frames per second". I know myself that the two access points I have - one B/G and one N - are sitting there most of the day when I'm either sleeping or out at work dealing almost exclusively in beacon and management packets and that the data is only going to be there whilst someone's in the house (everything that's usually on, except the Wii, is connected by cable).
As such, an apartment block with 30 access points and people out from 8am-7pm most days will generate a lot of management traffic with almost zero data. It would be interesting to know how the distributions changed throughout the day as I'd bet that there is far more data during, say 8pm-10pm, than the overall mean. I'd expect daytime usage to be even higher skewed towards management over data too.
Additionally, it'd be nice to know over what period they conducted the sample - was the kit left running for a month or did someone just test it for an hour during the day..?
Actually microwave ovens are fairly interference-poor by design. The magnetron dumps the energy into a faraday cage which more or less keeps all the radiation inside (that's the point, otherwise it'd be useless at cooking the food!).
So they are a cause of interference in the Wifi bands - but not as much of a one as other devices.
"As I doubt if the ovens are linked to a frequency standard (mine aren't) or are crystal-controlled (I've got to whip one of mine apart 'cos of a funny sound - pictures at 11 - to discover this fascinating fact), they might stray into the WiFi band. Or just block the WiFi receiver."
As a point of information, they are in fact intimately linked. The 2.45GHz band is one of 11 internationally agreed Industrial, Scientific & Medical (ISM) bands. These bands are unlicensed because of the acknowledgement that various ISM devices such as microwaves, MRI scanners etc are likely to blast out noise in that band, making them impossible to regulate. Microwaves *have* to operate in the 2.45GHz band because that is the wavelength which produces the correct excitation in water molecules which is their whole principal of operation. 2.45GHz was deregulated *because* of microwaves - it's not a coincidence.
Microwaves aren't actually supposed to blast out noise, but they will if their shielding breaks - which is does occasionally. Licensing that part of the spectrum to a company would be a little unfair because every so often a broken microwave would take out their communications. Instead, OFCOM has allowed a free-for-all and we've developed noise-resistant devices such as Wi-Fi transceivers which can, in theory, operate in the presence of interference.
Hope that's useful.
Paul
http://en.wikipedia.org/wiki/ISM_band
http://www.ofcom.org.uk/radiocomms/isu/ukfat/
Blimey. That must be some awesome super high-fidelity, surround sound baby listener. 4kHz of spectrum ought to be enough to convey a voice channel, although I must admit I've never tried to hear what a screaming baby sounds like on a communications channel. Lets say 50kHz for good measure, including guard bands. 8MHz could carry a video channel, in fact probably 2, just about. There's either some very badly designed baby listeners or some confused reporting here.
Why, out of interest, are the baby monitoring devices using so much bandwidth? You could quite easily fit 180 cd quality channels into that space and despite what people may think, their childs voice does not need to be rendered at that quality.
Is it just sloppy design on behalf of the manufacturers, or is it just cheaper to use components with less tolerance and wider filters?
Thumbs down for using 8000kHz when 10Khz would do nicely.
It's true, I gave up with my (admitedly cheap) tv sender as neither it or the wireless could cope with each others interference, even after trying different combinations of channels.
As for microwaves while in theory they do interfere in practice they are generally reasonably sheilded and more to the point not used for more than a few minutes a day - in fact if these surveys were done during the daytime then there is probably a very good chance that none were used within the range of the detectors.
A couple of years ago, I spent a couple of hours trying to get a usually reliable 802.11g connection operable for a laptop in an urban office.
I had tried all the usual tricks and was ready to pronounce the unit 'dead for unknown reasons'.
I was waiting in reception several floors for the lovely office managerette to finish a call to say it was no-go, when I noticed a 20+ car Wankercade/Cuntourage of black limo$, black explorers, motorbikes etc leaving.
It was Wanker(c)Google - Blairs' lot leaving from the Wangham Hotel in an un-named antipodean North Island city, after some big conf.
Then someone at the back of the office shouted with glee - ITS CONNECTED FINALLY
Wankers' spooks must have been jamming the spectrum I say - fucken arrogant cocks.
Black Helicopters - to match the motorcade.
If you buy a router it has WiFi whether or not you want it. If you have a desktop PC, it's probably wired to the router as few desktop machines seem to include a WiFi card.
WiFi can be turned off it not wanted, but how many owners know or care? I imagine there must be a few million routers emitting unnecessary WiFi. We're constantly told to encrypt it, but not to turn if off it not needed.
Friend replaced his microwave when he realised it stopped his DECT phone from working. The replacement was okay. My WiFi won't work if placed 'close' to my DECT phone, I'm pretty sure I'm not the only person to have found this sort of interferance out, presumably from hardware which isn't quite as 'accurate' with its' frequency. So why has it taken ofcom so long to find out? Surely one of the reasons of ofcom is to ensure there isn't multiple uses of similar frequency's that makes useful devices into useless ones?
oh and my new video sender seems to not like being near the phone, that or I'll have to put up with the rubbish signal.. Still , it was only 20quid ,so can't be too unhappy.
Just glad I only use the WiFi for the Wii, wired networks all the way..
just walking down any city street and you will pick up at least a dozen wifi AP's within the distance of 200 yards.
though some AP's seem to have problems with EHS sufferers (who just happen to dislike thier neighbours enough) that suffering the thud thud thud effect of a wifi jammer is better than the constant howling scream of a AP talking to a laptop in the next flat.
linky's
the strawberry jam: http://www.tayx.co.uk/wbj01-wifi-bluetooth-jammer.html
what the physical effects of exposure to that AP/Crack berry will do to you: http://www.wave-guide.org/library/studies.html
an RF detector that works like a bat detector ie gives a audio feedback of the rf recieved: http://www.emfields.org/equipment/a-com.asp
(would help if they started making the damned things in bulk, im sure everyone should have one)
heres what i can feel (audio track) from nect door:http://www.emfields.org/equipment/audio/wifi.mp3
and the noise from the 3G tower 200+m away: http://www.emfields.org/equipment/audio/3G.mp3
combine those together and the idiot nearby with the DECT: http://www.emfields.org/equipment/audio/dect.mp3
and you have one very unhappy bunny, that is looking for somewhere very quiet....
hmmm i wonder... is there any where on the russian step(mongolia) or in the Gobi that would be safe from this radiation....
(i'll put up with a bit of ionizing (a/b~g/x)radiation anyday to this insideous non-ionizing (a-b-g-n-t) rubbish)
Mines the one with heavy Tin-Foil for a lining......
which also works well as insulation for the flames in the future revolution where the corrupt politicos and telecoms bosses are fed to the fire........(for thier crimes against humanity and the planet)
btw.. if anyone does know of a location in the UK that is utterly mobile phone/AP radiation free, do let us know, there are an awful lot of EHS sufferers looking for a safe bolt hole to escape to...
A picture says a thousand words, so here's my picture:
http://chezphil.org/wispy_for_hire/
That's what I found when I used a spectrum analyser to work out why my WiFi had stopped working. The girt big streak down the right hand side was clearly the problem. Guess what - the couple next door have had a baby...
@Andus: microwave ovens don't seem to be much of a problem, as far as I can see on the WiSpy. And they're not normally on for long - unless you live above some dubious fast food establishment.
umm the average microwave oven is shielded, all of the magnetron's output is fed to the interior of the oven. Any leakage of the RF outside is considered to be a Bad Thing.
And I wonder if Ofcom picked up any Amateurs boosting their WiFi with high gain antennas in the 2.4 to 2.45 GHz band!
I've had a look at the 2.4GHz band around my house. I can see the baby monitor in my house and what turns out the be the baby monitor next door. We have a small intermittent spike that I haven't managed to track down, but I think it may be a doorbell. None of this has any negative impact on Wifi performance. However what I can see is 12 other APs that do make it hard to get a good signal. When it boils down to it Wifi is problematic because there are really only three clear channels that don't interfere with each other. In a location like an appartment the odds of getting one of those channels free of interference is impossible.
Yes in a busy area other things will interfere, but the most widespread use of the 2.4GHz band is wifi and therefore the thing most likely to be interfering with wifi is more wifi. The only sensible solition is to go to 11a.
@Andus McCoatover
I have heard tell of microwave ovens affecting wifi, but I've never encountered it myself. Given the power output of a microwave oven I wouldn't want to stand near one that was leaking enough radiation to interfere with the wifi networks in neighbouring houses.
Why does a baby listener require 8MHz of bandwidth when you only need transmit audio frequencies of 8KHz in range? Now I know they might be FM modulated and that requires more bandwidth than AM which would 2x8KHz but 8MHz seems excessive.
Poor filtering perhaps?
The article seems to suggest that digital transmission will be better..not if like for like! It will be worse! Unless they resort to data compression and/or reducing the audio bandwidth, and changing the RF modulation.
I could have told them that -- five years ago. There's no magic to wireless but judging by the way that people talk about it you'd never guess it.
Now, having figured out that WiFi's being held back by being crammed into a tiny slice of spectrum that nobody wanted because they deemed it unusable maybe they'll free up another tiny slice or two somewhere else in the spectrum. Its public property, after all.
The bit about microwave ovens is important. The reason why the ISM (2.4GHz) band was deemed unusable was because its a frequency that's absorbed by water (that's why microwave ovens work....). Now, if you delve into the design of the MAC protocol for 802.11 you'll find that it was designed around the idea that microwave ovens pulse their radiation at half the mains frequency (I don't think they do now, BTW) -- the idea being we could get a frame in during the off cycles. As for the frequency being crystal controlled, its not; the frequency is generated by a cavity magnetron where the frequency is dictated by the size of a slot machined in a piece of metal.
"The paradox of the Commons".
Not everyone is "nice", so the bad people "win". Oh, well. On to 802.11n. Up to 5GHz.
On microwave ovens: They are SUPPOSED to contain their emissions. Most do or some regulatory agency would be like white on rice, not to mention the government...
"Not that bad Wi-Fi is always to blame - Mass found that users are quick to blame the wireless part of their connectivity when the fault actually lies in ADSL, routing or even the servers at the other end of the connection"
I lived in an apartment where WIFI was the only way to get data from one side of the house to the other. WWW/Voip/etc would occasionally/randomly experience 100% traffic loss (1 - 30 minutes at a time). I bought new gear, range extender, etc in vein to try and make it work. I never trusted the WIFI link and was always fiddling with it to fix our internet. It wasn't until we moved out that we discovered the fault was within Optimum Online's cable network. Turns out that's it's a common problem on their network even with "full strength" at the cable box, the neighbors suffered the same problems. My mistrust of WIFI made me blind to that possibility that the daily service outages were from the cable company.
I previously worked for a company that had 2 small remote sites, one had a direct connection to our main site and the other used a wifi connection to the connected site (they were across a very small lane to each other, the wifi connected site being only about 3 users).
Every dinner time the wifi site would drop off the network, it took 3 visits from different techies until I went and happened to realise that it was when someone turned on the Microwave to cook their dinner in the cabled site that the wifi site dropped... it took a few attempts but we eventually got it working (and asked them to contact IT Services if anyone ever had a bright idea to change the Microwave for a new one).
- That said I've also seen a network drop every dinner because a tea lady routinely parked her tea trolley on a 10BASE2 ring networks cabling.
It's an unregulated* band so if neighbours are interfering, too bad. Sure it's annoying if it doesn't work, but that's the way it is.
One of the benefits of living in country towns is that WiFi spectrum is fairly well clear. From my house in a country town in NSW, Australia, I can pick up 4 access points. 3 are mine, and one is the school across the road.
So to all those who complain their wifi is too slow... Shut up, plug in a cable, get on with life.
* well, not true - there are regulations still on power output, etc
Ok, not exactly a myth but misunderstood.
All WiFi devices recieve over the whole WiFi spectrum but the signal gets weaker the further away from your channel it gets. Pretty much the same as increasing the distance or number of walls between stations.
In WDS mode you have to set the channel on both stations manually. Yes it works best if they are both set to the same channel but it still works if they are set to different channels, providing they are close enough.
WiFi devices themselves actually work quite well together but not everything using the 2.4GHz spectrum is careful not to trample another signal.
The chipset of the device is very important too. Atheros devices generally work a lot better than Broadcom in my experience. The Broadcom ones can get so bogged down with other peoples noise that they don't work. The Atheros ones seem to be able to use the frequencies that are operating.
WiFi is similar to ADSL in that it transmits on lots of little channels at the same time, some of which will fail and some will work.
Eventually people will stop using the old baby monitors as their children get bigger. They will also get fedup with video senders and get a cable put in. People should also put in CAT5.
The ISPs issue wireless routers routinely and people are too stupid to realise that a wired connection is faster and more reliable.
I've seen setups where a desktop is directly next to a wireless router in a bedsit -- not even a laptop as an excuse for wireless.
In my neighbourhood full of studio flats you could barely swing a cat in, I can detect an average of 33 wireless routers -- most on default channel 11 (which shows how little they know about changing channels).
My immediate neighbours' router was on the same channel as me about 15 feet from my computer, badly interfering with reception from our router which is downstairs and about 50 feet away.
Fortunately my neighbours are some cute Scottish girlies and their router is on auto select. I could make it switch channels simply by placing a couple of my redundant wireless routers next to the party wall and asking one of the girls to cycle the mains to their router. But such cooperation is rare, given the high turnover of tenants (thanks to so-called "Assured" Shorthold tenancies leading to people moving annually).
Take a look at the door on your microwave oven, and you'll see a mesh inside the glass. This is a small Faraday cage to prevent leakage of microwave radiation from the unit.
Want to see how effective it is? Take your laptop and wireless access point to your microwave oven, unplug the oven (don't want to nuke your electronics), and slip the access point inside. Shut the door as much as you can without damaging the cable powering the access point. You just try any get a connection to that access point.
You can try the same thing by again unplugging your microwave, and placing your mobile phone inside and shutting the door. The signal strength will drop to zero.
Anything on 2.4 will interfere and thats a huge range of things . Add the fact that wifi abhors walls , floors etc and most users have no clue how to set it up and its a pretty good lame duck .
Id rather use cable and just not worry about it . Homeplug is a great alternative but again that has its problems .
People looking for any form of quick fix are going to be disappointed . Radio frequencies are finite and whatever standard its going to break down as soon as it gets popular .
Given that this is unregulated space you could equally make the argument that all you selfish WiFi users are causing interference to the rest of the population who are trying to watch telly in bed whilst listening to the little ones. I gave up with my video transmitter because the interference was just soo bad. You should all move onto another band and let the rest of us carry on in domestic bliss. :)
I've also given up on WiFi now and we stick to Homeplug for our home networking and ADSL sharing needs. Much faster, more reliable, lower latency etc etc.
I'm sure many others will point out the same thing: a Magnetron is a tuned circuit. A crystal wouldn't help to stabilise the frequency. But quite rightly, the frequency is not very accurate as it relies on physical dimensions and, therefore, manufacturing tolerances. However I don't think it would vary 0.5GHz from its nominal frequency, so shouldn't affect WiFi, and Microwave ovens are shielded to a high degree to prevent them cooking the cook's innards, so there shouldn't be much radiation to interfere anyway.
Folks, I guess some of you don't have an RF background. I do. 34 years in radio, effing big water-cooled bottles to mmic's. So I know a tad or two about wireless, er, I mean radio.
So, microwave oven remix.
Point is, you can't attenuate anything totally. It's asymptotic. (Whew! Haven't heard that word since school...)
eg - @Ash - my phone worked in the (unpowered) microwave - I tried it - got Her Indoors to ring it. Why? 'cos there's a fuc*king big basestation across the road. Phone's sensitivity is maybe -110dBm, and that mother shouts at +46 at a guess. Not even counting the antenna gain.
(Plus, the holes are designed to attenuate 2.5G, NOT 900 smeg GSM, but that's incidental)
Nothing's perfect. If I get 60dB attenuation from that €60 microwave oven, I'd be gobsmacked. Especially if it says "Made in China" on the side. So blocking of the WiFi front-end is quite probable. A similar "Made in China" sticker on the router does nothing for my confidence.
With apologies to Huawei.
Oh, yes. In memory of my behaviour watching Dr. Who and the Daleks, I've hidden my router behind the sofa. I reckon that'll help.
Sorry that's a bit of a myth, the larger percentage of the population is actually early twenties to mid thirties. With a large portion of that group actually being single and living on their own. From what I can tell the online dating business is big in Bournemouth as well.
The students figure is probably right though, there are a lot of langauge schools in the area besides the Uni.
(Note to self, get baby monitor that uses WiFi network rather than kills it).
A lot of the hardware on ISM bands is regulated, if only for the assurance that it works on the correct frequency. The key point about ISM is that it doesn't require individual licensing. So I can walk into Tesco and buy a radio transmitter. (Bluetooth ear-thingy).
Changing the aerial of a Wi-Fi unit is very likely illegal. Even though it might reduce the RF Energy going in un-wanted directions. You or I might be able to set up something which doesn't harm other users, but knowing some other users... Can we trust everyone to be the right sort of clever-dick?
Seems like everyone has suddenly woken up to what happens when millions compete for a limited resource in an unregulated market.
The real question is; what took people so long to wake up ? This was entirely predictable from the outset.
As for, "move to 5GHz and you problems are over"; for how long ? Those people are suffering the same delusions that 2.4GHz would be a panacea to all their problems. The same is true of those who suggest bumping up power output; only works so long as you're ahead of the curve and every man and dog isn't doing the same.
How does it go ... if you don't learn the lessons of history ...
Mine's the one with the CB radio in the pocket.
" Anyone relying on 2.4GHz 'free' access is asking for trouble, fine for limited home use but not for important services. If there are any professional idiots out there using 2.4GHz for industrial control or similar potentially safety related projects, have you considered how easy it would be to modify a microwave oven to saturate your whole factory's RF links and do DoS attack?"
Actually, if you ask people in the industrial control business the one thing they say is never use wifi for mission critical stuff.. eg DNC from a PC to a machine tool, too mush risk of a missed/corrupted packet which can result in a serious accident
I'm afriad us industrial people are stuck with miles of network cable everywhere :+(........ and even that has to be correctly shielded and grounded
I live alone so, barring avoiding the hazard of trailing wires, what would wireless do for me? It's the most overhyped thing going. Because it's comparatively easy to produce, sell etc, it seems to me the consumer has been unthinkingly led to it. If you're a busy household running a lot of connections at once, great. If you're a sad old billy-no-mates like me, using one machine at a time, it's like paying for a hotel's worth of facilities when you only need a room.
I gave up trying to run WiFi in my house yonk ago, the signals would work fine for about a week, then degrade to such levels that I couldn't even get connected to my AP no more than 20 feet away. Changed channels, even bought a stronger AP with bigger ariels. In the end I simply took the plunge and cabled my house in cheap CAT5 and now no problems. I still run my WiFi, just in case I want to whip up the laptop, but most of the time it barely has enough juice to watch a naff YouTube vid.
Tin foil is cheap and easy to apply behind wallpaper & under carpet, its tricky on the ceiling, without reboarding and artexing afterwards but when properly connected it does ensure a good nights sleep...
Remeber there is no need to hate your neighbours because of your problems. They are not resposible for your EHS afterall.
idiot down my road hsa got themselves a cheapo belkin 802.11n AP and is running it in the 2.4GHz spectrum with channels 1 and 6 bonded in 40MHz mode - they've sucessfully messed up the 5 APs that were running on channel 1 (3 of them) and channel 6 (2 of them). there were already 3 APs in channel 11 so that mgith get crowded.
of course, we live in the UK and so can use channel 13 too - but very few appliances and
devices are easily modified to allow this operation cos its mainly US kit - those Japanese get channel 14 to play with too.
whilst 11n could be the panacea, it wont be if everyone uses the 2.4GHz mode - its 5GHz all the way or live in the old era.
actually John 99% of wi-fi gear which includes routers and especially the Wii do NOT turn off wi-fi. Instead they disable the receive part of wi-fi (because they have lame programmers and it's easier to do so!)
The Wii for example uses the same output power when in standby mode as it does when enabled, but in it uses about 5 to 10% less when disabled. I had discussions with Nintendo last year about this and they refused to do anything about it.
While this wastes large amounts of electricity, people waste even more by leaving it on all the time, even when they are in bed, in work, or on holiday and it is not even needed!
As an alternative why not try Homeplug 200 adapters which can handle 2 simultaneous streams of HD video easily, are not affected by your neighbours and usually not other equipment either?
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2024
