Anti-social Wi-FiWi-Fi is governed by the IEEE 802.11 standards set. The best known of these are 802.11b with a maximum raw speed of 11 Mbps and 802.11g allowing a maximum of 54 Mbps both with a nominal indoor range of 38m.
The frequency range (2.400–2.485 GHz) is divided into 13 x 22 MHz channels but spaced only 5 MHz apart and attenuating to -30dB at their edges. Regulations only permit the first 11 to be used in the Americas and some European countries e.g. France and Spain, are even more restrictive, but most allow all 13.
Interference causing reduced throughput can be minimised by carefully choosing which channels to use, avoiding clashes with neighbouring networks and also outside interference from other users of the frequency band such as Bluetooth, cordless phones, microwave ovens and remote controls.
A consequence of the oversize bands compared to their separation is that neighbouring networks have to be 4 or 5 channels apart to avoid interfering with each other. For example, Channel 6 covers 2426–2448 MHz to -30dB which just touches the edges of Channel 2 which reaches up to 2428 MHz and Channel 10 which reaches down to 2446 MHz. With only 11 channels available, these limitations mean that in practice only channels 1, 6 and 11 are usable with a little adjustment available to avoid other strong broadcasters (though it would require specialised equipment or a lot of trial and error to achieve that).
In theory, in the UK, four useful bands could be achieved—1, 5, 9 and 13— but in practice too many people know about the US limitation and (some equipment comes preset) to get this to work without mutual local agreement. One of my devices only allows 1, 6 or 11 to be set in any case. As wireless networks get more popular, finding a free channel to use gets more difficult but, unless you are working in a very dense environment, it should still be possible to find a working combination.
There is a proposed new standard, 802.11n, which has not yet received approval (and won’t until at least the end of this year), and offers a raw speed of 248Mbps+ and a 70m indoor range. It achieves this by using a facility called Multiple Input Multiple Output (MIMO) which uses multiple aerials and a wider 40MHz bandwidth (channel bonding). This effectively uses most of the 2.4 GHz band in one fell swoop, blotting out everything else within range.
In theory, the channel bonding is optional and the standard requires the stations to detect activity on neighbouring frequencies and to moderate its use accordingly but new “pre standard” devices (called pre-N, draft-N, Extreme, RangeMax, Turbo, Super-g etc.) have appeared (technical article) and been heavily marketed for a couple of years which don’t necessarily conform to all (or any) of the features. For example, the draft standard allows for the use of the 5GHz frequency band, but in order to retain downward compatibility with b and g type devices, most of the equipment produced does not use it unless specially configured (if at all). They also tend to come pre-configured for maximum speed and the onus is on the user to change that, which is unlikely because that is why they paid for it in the first place. In many cases, users of this equipment have only one item (usually the router) to this standard with the remainder operating at the g level so the benefit is lost anyway. Devices from different manufacturers frequently do not interoperate, certainly not to their full specification.
But for must domestic purposes, why would you need this speed. The original type b system exceeds the speed of most broadband connections. If you have a local server or communicate locally between machines then the speed of the g system may be useful but for any bulk data transfer, like a backup for instance, it would be much better performed using a wired connection.
Note: The raw speed figures quoted for wireless transmission are absolutely worthless. Realistic figures, but under ideal conditions, are 6 Mbps for b, 23 Mbps for g and 90 Mbps for n.
Webmaster
OK – so you’ve made me feel guilty! I’ve just bought an Apple Airport Extreme Router. My feelings about it so far are:
1. The MIMO technology is the best part about the new draft-N standard: the range is much, much better. So let’s hear it for better antennas.
2. The router config is nice, once you’ve worked out how to get to it. The thing self-configures, but if you want to do something clever, you’re going to have to do something thinking on how to do it – but maybe that’s rather clever psychology.
It seems from your link that the key issue is going to be which chipset manufacturer is going to win on the debate over spectrum usage, ‘cos most off-the-shelf products (wifi routers/access points) are all using a limited number of chips to deliver the functionality.
There is also the point about upgrading to 5GHz: which the Apple product does *provided* it isn’t being asked to serve pre-N laptops in the vicinity – yeah, right …
It looks like the Intel chipset deliberately doesn’t allow the 40MHz channels on the 2.4GHz band so that is acting good-neighbourly. Perhaps the others will follow suit.
I had a look at the Airport Extreme and one review said that 5GHz wasn’t available in the UK. I hadn’t spotted this before and am not sure it is right – surely 802.11a is available and it uses the same bands. Also, according to another review, the UK doesn’t allow 40MHz bandwidth at all. I don’t think that has stopped people selling other pre-n and turbo routers. It looks like with this device you are not being anti-social at all.
What they seem to have got wrong is only including one radio. The proper way to do class n wi-fi is dual simultaneous radios in the base station and a single switchable frequency radio in the mobiles. Then you can always get an optimum connection anywhere.