Blame baby monitors, not congestion, for your WiFi woes | Hacker News Reader

Blame baby monitors, not congestion, for your WiFi woes

arstechnica.com

32 pointsterpua15 comments

WiFi has become important enough—and the interference problem bad enough—that the report recommends creating a new "2.4GHz friendly" logo program to brand devices that are good radio-frequency citizens.

Are they kidding? Since 802.11_ protocols are 'smart' in the sense that wifi nodes exercise congestion avoidance with other nodes running the same protocol, the next wifi standard just needs to be allocated dedicated space in the spectrum, rather than having to share the same narrow common space with baby monitors and AV extenders. Alternatively, the band of less-regulated spectrum needs to be expanded wide enough that manufacturers can create some informal standards to avoid stepping on each others toes.

(And yes, trust me, I know why this is politically difficult, and that it wouldn't immediately help the enormous number of users with current-generation hardware. But really, their recommendation isn't likely to make much short-term difference either.)

In 2.4GHz band, there are only three non-overlapping channels (1,6,11). For the optimal use the wifi and baby monitors need to far apart.

Current wifi access points detects these interference and do not settle in a channel where baby monitors are operating. For better performances people should start moving to 5GHz band.

I've been a ham operator since I was 10 so I have a little experience on the subject. Interference can come from anything. Lots of power supplies for surprising devices are culprits. The little cheap universal AC/DC adapters are one of worst offenders. Power-lines, bad transformers, any business that has massive electrical devices.

TV's, (especially older) microwaves, vacuum cleaners... pretty much anything can cause interference. One of the worst culprits are the cordless handsets.

The biggest offenders aren't devices that use the same rough frequencies, it's ones that either splatter all over the place or have harmonic issues (multiples of frequencies, a 500MHz device can splatter 1,000MHZ, 1.5GHz, etc)

It would be nice to have more spectrum for use for digital devices. Hopefully in the US they continue moving the TV stations off and eventually provides more spectrum for growing wireless devices.

Yes. At home, I had to muck with the channel on our router/access point so our cordless handset wouldn't interfere with it. Nothing worse than losing network connectivity because the phone rings...

Not all baby monitors are at fault for WiFi interference. In the US you can look up the product's FCCID in the FCC database and it will tell you what frequency it uses.

http://www.fcc.gov/oet/ea/fccid/

I'm not sure if the UK has a similar licensing system search or not.

Wouldn't making baby monitors that use Wifi mitigate the issue?

That would drastically increase the cost of baby monitors. Right now they use very simple analog RF techniques. Switching to 802.11 would increase the design complexity and cost.

wifi access points sense the wireless medium around them before transmitting. If they find it busy, they do an exponential wait, and try again. Thats why everyone can be on channel 11 and things won't be too bad. If they could sense the medium perfectly, routers would always be able to guarantee service regardless of the network congestion just throughput would decrease. However, they can't sense the medium perfectly, so we run into the hidden terminal problem, where wireless packets clobber mid route because they're transmitted by routers far way from each other enough that they can't sense each other. Thats the reason why wifi congestion stops traffic in congested parts of london.

Baby monitors/any-analogue device doesn't do this, and that's where the problem comes from. The 5GHZ band (802.11n) is a good alternative.

http://en.wikipedia.org/wiki/Hidden_terminal_problem

There's a great low-cost (sort of) spectrum analyzer called "Wi-Spy" (no affiliation) for anyone who wants to track down what's ruining the 2.4GHz spectrum. With that and WireShark running on a Linux machine with monitor-mode capable Wi-Fi interface you can pinpoint what your Wi-Fi problem is quite well.

>> "We have found that it is rare for the user data frame rate to exceed 10 percent of the total frame rate,"

Do not know how they came to this conclusion, beacon frames are the only periodic management frame.

There are lots of probes and probe responses in addition to the beacons. If you have CTS/RTS enabled (most don't) then you have those frames as well.

Each packet has to be ACKed by the receiver so there's lots of those.

Management frames are usually short, though.

Agreed, the probe requests and probe responses do not hog the medium. As noted most of us do not enable RTS/CTS, it would be clear if the researches pointed out the exact conditions under which they performed their tests. With out which the statement has no weight.

We haev free Wifi in our cafe bar, so do the other 3 cafe bars on our street. All four use Ch. 11! I told our techs but they can't work out how to change it sigh.

On a busy day, stand on the bar and yell "Free latte to the first geek who can fix our wireless!" It'll be sorted in five minutes.

Are you sure you can't guess the wifi device admin password and fix it yourself? (You might be surprised.)

Reply on news.ycombinator.com