[asterisk-biz] VoIP 9-1-1 failure - don't let it happen to you

[SIP]

Peter Beckman wrote:
> On Tue, 6 May 2008, SIP wrote:
>
>   
>> Neither of which are terribly effective. GPS is rather unreliable
>> indoors, and if you're truly on the move, having your phone stop working
>> every time you switch APs until you reconfirm your location (which you
>> may not even know) is, perhaps, a worthless solution.
>>     
>
>   The ISPs could provide IP-to-Address mapping for dynamic IP assignments.
>   The ISP knows who the customer is (at least in theory, technically) when
>   they dole out an IP to them, and if the service is something like FIOS,
>   DSL or another physical entity, the ISP should be able to provide
>   registered ISPs a service where the IP can be queried for the address, and
>   potentially matched to the data provided, or maybe even as a trusted
>   source of the data.
>
>   While technically pheasable, and with most of the ISPs having been phone
>   companies both currently and in a past life, and their knowledge and
>   experience with providing 911 services in the past where people didn't
>   have to tell the phone company where they were, it would probably take the
>   Government to take action to force the ISPs to do this.  And even then, I
>   know there are people on the list capable of shooting holes in this as
>   well.
>
>   I do like the GPS idea, but make it aGPS.  You don't need to have service
>   with a wireless carrier in order to listen to the signal or talk to the
>   tower.  If you know where the tower is, you can get a decent idea where
>   you are +/- 1 mile and push the call to the right 911 center.  But then
>   again, sure, if you are in the boonies with no cell towers, you are hosed.
>   But if the ATA can pick up a signal, and it DOES get a GPS fix, it should
>   be able to pass that or be query-able.
>
>   
>> These are basic problems for which NO one has a solution.
>>     
>
>   That's because there is no good one-size-fits-all solution.  What I've
>   thrown out is just more of a "it might work in some cases" solution.  This
>   is a complex problem, and nobody has an easy solution.  I think the best
>   solution will be a combination of technical solutions and good testing.
>   There really should be a new number, "912" that routes like 911 except to
>   a recording that states the location you would reach if you dialed 911,
>   and the information they have on file (or passed in the call).  Easy to
>   verify, easy (well, maybe) to implement, and no operators annoyed with
>   "test" calls.
>
> Beckman
> ---------------------------------------------------------------------------
> Peter Beckman                                                  Internet Guy
> beckman at angryox.com                                 http://www.angryox.com/
> ---------------------------------------------------------------------------
>   

Right... that's mostly what I meant by no one has a solution and it 
would require architecture changes. One could assign a physical location 
to every access point, and be required through some trickery of 
currently non-existent protocols to pass that information onto the PSAP. 
But then you have the privacy people worrying about whether or not this 
constant tracking information could be obtained and used by stalkers. 
And once you bring a technology like WiMax into play with a theoretical 
(and I am aware of its current limits in practical installations) range 
of 30 miles, then what? A 60-mile diameter circle is a big area in which 
to get lost without any possibility for triangulation. At that point, 
you'd have to rely on GPS (if not in a city with high buildings or a 
wooded area filled with trees), which adds in yet another thing that 
could go wrong.

This is the real issue with VoIP and E911 -- the infrastructure. 
Wireline phone is easy for E911. It stays in one place. Its circuit is a 
known location.  Cell towers have a limited range, can triangulate if 
necessary (for scenarios in which GPS wouldn't work), and when outdoors, 
you have the GPS option. And there are cell towers EVERYwhere.  IP, 
however, is nebulous from its very architectural design. It was built 
that way intentionally. This makes for protocols which travel over IP to 
be somewhat nebulous by nature.

Unless we radically change the variables involved, I'm afraid it will 
only get MORE nebulous, not less so. We may solve one problem now, but 
be faced with another in 10 years and another 5 years after that, etc, etc.

N.