To many, the E911 issue caused by VoIP is a growing problem and people should be concerned. Presently the E911 solution requires too much honesty on the part of the user, validation on the part of the operator, the intervention of a third party to relay the data to the right PSAP and a lot of prayer on the part of the person who needs it most.
But in my view VoIP is being unfairly targeted when in reality our entire 911 system in the USA needs an overhaul. Have you ever called 911 from your car in California after observing an accident, seeing a stranded motorist or worse a staggering homeless person walking in and out of traffic lanes on a 65 MPH freeway? Chances are you can be five to ten miles down the road before you get to speak with a live operator.
With VOIP, users can pick any area code and make that their VoIP numbers. Gone is the day where geography ruled the roost and where you were determined your phone number. Now a 212 number can be in San Francisco, a 213 in Boston and a 214 in Anchorage. Hardly where they are best know for being from. This nomadic VoIP situation is causing the E911 services fits and making life a living hell for companies like AT&T, Earthlink and AOL who are all working hard to not run afoul of the laws.
Some ideas on solving 911:
GPS chips in the routers - I know this is very 1984ish, but there would be very little question where the call came from (unless someone is using a VPN)
1-Area Code + 911--> This requires some smarts on the part of the caller but it would automatically route the call to a PSAP (Public Service Access Point) close to the location.
Do you have other ideas?
To me, E911 is a problem for VoIP not only because of flat addressing of IP but also the way all the current VoIP protocols are designed. None of them concentrated on E911 issue beforehand. They came out with a protocol and accordingly designed their network. Thus getting any help from them is difficult. Of course, IP behavior is the main obstacle for E911.
This SOA based VoIP architecture, Service Oriented VoIP (SOVoIP)
http://jubaerarif.blogspot.com/2007/04/sovoip-new-voip-architecture.html
focusing on E911 from the scratch. This architecture is still long way to go but certainly a good start!
Posted by: Arif | June 11, 2007 at 09:38 PM
The only real way to locate a 911 caller is to have technology in the phone adapter, laptop or other end device that can be located when the user dials those 9-1-1 digits. The cell phone industry has done this for years.
There is a good blog about this at http://www.e911blog.com
Posted by: Ken Arneson | June 06, 2007 at 11:48 AM
How about using RSVP to send a query from the PSTN PoP to the router nearest to the user ? This is like the GPS solution but would require no changes to existing hardware. There is flexibility in the packet format to allow this. See (http://www.cisco.com/univercd/cc/td/doc/cisintwk/ito_doc/rsvp.htm) for reference.
Posted by: Tony Palik | June 06, 2007 at 06:14 AM
I have made the following point earlier: the fundamental problem is in the tacit assumption that E911 service is a “voice” service and then we are trying to extend it to VoIP. Instead we should consider E911 service is part of “access to PSTN” service (after all E911 is available in the Fax line as well) and so we should extend it to “Internet access” service. If we do that, then as you observe, wireline access providers can offer location service to PSAPs. We need to solve wireless access providers, just like we had to do in the wireless PSTN.
There is an additional benefit in doing this - as end points and PSAPs evolve, E911 can be more than just “voice” application. One can connect medical monitoring devices , for example.
Posted by: Aswath | June 05, 2007 at 08:08 AM