There's been a number of scare stories about the risks of mobile phone masts cooking small children and the like in the media. Whilst there's no conclusive evidence of any risk and some simple maths and physics shows that masts are probably less of a real risk than handheld phones, there is still not much sense in phone networks installing antennae virtually inside someone's house, what were they thinking?

Thanks to Petri for spotting this lunacy. There are far better ways of installing antennae, the Undetectables for instance do some particularly cunning installations, and Orange have some masts that look like trees.

Can you spot the masts in these pictures?

Quick overview of how radiation levels diminish over distance

Radiation exposure diminishes with distance on a square law relationship, i.e. if you double the distance you quarter the exposure. Hence a 1W mobile phone 10mm from your brain (10,000W/m2) exposes you to 10,000 times the rediation that a 100W mast could generate at 10m (1W/m2)! I'm guessing a little at power outputs whilst 1W is a typical value for a GSM phone, also I'd expect a base station to output far less than 100W, maybe as little as 5-10W.

This doesn't take into account duty cycles, where a base station would transmit almost continuously whilst a handset would transmit as little as possible, but it still shows that a handset will expose you to higher radiation levels than a base station ever could.

There's currently a sticking point with how one should represent a Bluetooth device in one's Foaf file from a perspective of both privacy and spoof-a-bility. A Bluetooth device effectively has two id's; one is a user-changeable text field typically "Jims Phone" or more frequently a factory default like "Nokia 3650" and the other id is a 12 digit hex identifier not dissimilar in concept to a network card's MAC address.

Both of these are modifiable (although the hex id is rarely changeable in a mobile phone) so spoofing is always a posibility. One solution that would make spoofing harder would be if the Foaf file was to contain a hash or checksum of both ids, this would be hard to replicate unless one had seen both the ids, but this would make generating the entry in the Foaf file entry more difficult.

Privacy is going to be tricky, although there's always ways of mitigating this:

• Don't put any Bluetooth information in your Foaf
• Restrict your Foaf to a trusted community
• Don't turn your Bluetooth on

Another thought - maybe it would be useful to put something easily identifiable in the user modifiable field primarily as an advert for FoafMobile, but maybe also to help FoafMobile apps know that this Bluetooth ID really is worth looking up. I'm proposing a user identifier something like "FoafMob Jims Mobile" actually how about "FoaM"? Short, sweet, and it should be easy to type with T9 text entry.