RE: [Geopriv] getting position from WiFi

From: Dawson, Martin ^lt;>
Date: Fri Oct 26 2007 - 23:53:20 EDT

Yes - there are a number of WiFi positioning solutions already implemented. Check out names like Ekahau, Skyhook, and Place Lab. In fact, do a Google on WiFi triangulation and you'll find lots of prior art. From an architectural perspective, the specific method used to do the location determination is a second order detail. A LIS in a WiFi access network will invoke whichever flavour it is equipped with. The HELD semantics for requesting location remain the same regardless. Of course, if technology specific device-based measurements are part of the plot, then the LIS would like to be able to invoke that measurement capability as well. That is why the device-capability extensions draft has been created. Note that GPS is an overlay network specifically designed for positioning; fully synchronized, with atomic clocks, and long code sequences explicitly designed for accurate ranging. Using communications networks for triangulation inevitably involves a compromise since the access is designed for doing something else; not for location determination. There are some terrestrial overlay networks designed to provide location functionality - see S5 Networks for example. Nevertheless, the good thing about using a communication network when location is required in a communication context is that you can guarantee that the network is where you need it. If there's no communication then working out location is the least of your worries. Depending on a separate network for location information creates a serial reliability problem - which is exactly the issue when it comes to relying on GPS in a communication application scenario. Sometimes it just doesn't work when your communication network does - and when your communication application needs location information. So - there's definitely good utility in being able to use WiFi for positioning. Hopefully, one day, widely adopted communication network standards will include specific signalling facilities and procedures purely for positioning purposes instead of having to make do with measurements of signals that are just there for establishing communication and carrying content. Cheers, Martin -----Original Message----- From: Andrew Daviel [] Sent: Saturday, 27 October 2007 11:50 AM To: Subject: [Geopriv] getting position from WiFi I just had this idea while listening to someone talk about siting wireless access points : - would it be possible to get a location for WiFi enabled mobile devices by triangulating off 802.11 access points ? Theoretically, it could work like GPS. You could derive an accurate location from an "ephermeris" of AP locations plus time-of-flight data. In practice, I think a clock accurate to 10ns in each AP to get 10m resolution might be somewhat expensive, and NTP won't cut it, you'd need a hardware solution. Just using the AP identity itself would give reasonable accuracy with normal (not high-power) access points. RFC 3825 alludes to this possibility, suggesting using DHCP to distribute a list of AP locations. This technique would work indoors, while GPS and to some extent cellphone service do not. (cellphones won't work here; we have metal-clad buildings with wire in the window glass and the towers are too far away). (The newer generations of mobile phone have combination WiFi/3G/cellphone connectivity and can in some cases allow a mobile Web client to roam seamlessly from WiFi to cellular. Future models may allow VoIP roaming too, so there may be some incentive to have positioning work in shielded locations) -- Andrew Daviel, TRIUMF, Canada _______________________________________________ Geopriv mailing list ------------------------------------------------------------------------------------------------ This message is for the designated recipient only and may contain privileged, proprietary, or otherwise private information. If you have received it in error, please notify the sender immediately and delete the original. Any unauthorized use of this email is prohibited. ------------------------------------------------------------------------------------------------ [mf2]

Geopriv mailing list
Received on Fri, 26 Oct 2007 22:53:20 -0500

This archive was generated by hypermail 2.1.8 : Fri Oct 26 2007 - 23:55:41 EDT