how-to block ads
|
jeffpulver
Premium,VIP
join:2003-02-08
Melville, NY
|  911/when Service Provider does not provide access
Hi All,
I’m working on my talking points for a free event sponsored by the Cato Institute in on September 9th – see:
»www.cato.org/events/030909pf.html
With regard to 911 requirements, if the service provider is not also the access provider, how can that company be responsible for a 911 solution at all times?
With regard to the growing number of voice over broadband service providers, “voice” in reality is treated as an agnostic IP based application which runs across the public internet and the local access of a high-speed service provider chosen by the subscriber. When/if the underlying local access network fails, how can the application service provider become responsible for 911 compliance?
Is it not possible that the rules for 911 need to be re-addressed for IP based service providers? Until the technology platforms are reflected in the regulatory platforms, is it not also possible that there will always be an uphill battle if one is fighting with legacy players?
Driving from Montreal to New York City during the day yesterday, after being stranded in Montreal due to the blackout, and experiencing first hand the huge outages of the wireless coverage, if there was an emergency to be reported on the roads yesterday, it most likely could not have been reported using wireless. Does this mean that any Wireless service provider would have been held responsible for such issues? What happens if I rely on WiFi as my means for accessing the internet and during a blackout the access point is no longer available and because of this I lose my ability to place calls on the Net? How can the application service provider be responsible for some aspects of the 911 issues if they are not in control or supply the access? And if there are exceptions, who decides the circumstances when the rules can be broken? | |
clecrupt9
join:2002-01-22
GA
| Those are some good questions!
I think they do need new policy, telecom is an absolute mess in the United States. I see the point concerning wireless and perhaps 911 standards need to be looked at and clearly written so end users, and service providers alike know what to provide and to expect.
I dont have the answers, but perhaps the 911 centers themselves need IP integration to help broadband phone come of age. Also maybe the broadband phone service provider could have real live 911 operators that took the 911 call and patched it through to the proper center wherever that may be.
When a provider doesnt have access is complicated, there is line from network failure and provider at fault. What are the current rules Jeff? If Bellsouth didnt route my 911 call right now, what action could I take?
[text was edited by author 2003-08-16 18:51:12] | |
jeffpulver
Premium,VIP
join:2003-02-08
Melville, NY
| quote:
When a provider doesnt have access is complicated, there is line from network failure and provider at fault. What are the current rules Jeff? If Bellsouth didnt route my 911 call right now, what action could I take?
I'm not 100% sure of the answer but I hope to find out the answer over the next few days. | |
clecrupt9
join:2002-01-22
GA
| Ok, I'll see if I can dig up anything too.
I think that forcing Vonage to be a full Clec like what Minnesota has done isnt the right move. But I dont think that 911 should be just left to whatever goes is fine.
Vonage has offered a 911 service and though it has problems, its a start. But I could today start a broadband telephone service and say I have 911. What exactly am I required to do? If I am a Clec/Ilec I must have trunking and facilities to manage 911.
In theory Vonage could do whatever it wants; offer 911, not offer 911, say it offers 911, offer its own variant, ect. I think thats scary when you figure all the companies that will soon join this space, some of them may not have the social fabric to care about it working.
In the end I hope that the Government says as long as people know that broadband phone 911 is best effort its fine, or I hope to see an innovative solution where it works, if not better than 911 currently does on the pstn. The more problems IP can solve, the better and faster it will replace the pstn. | |
WhyADuck
Premium
join:2003-03-05
| reply to jeffpulver
I think the first hurdle will be to get 911 providers to accept the fact that VoIP exists, that it's something they must learn to deal with, and that it has limitations.
A few random thoughts:
1) Every 911 center should have a reliable broadband Internet connection (maybe more than one if possible, from different providers, for the sake of redundancy). If choices are available, this connection should be obtained in such a way that it bypasses the local telephone company, so that a disaster at the local phone company won't affect the Internet connection. In practice this would probably mean using cable modem service, where it is available.
2) Every 911 center should have a goal of being able to accept 911 calls directly off of a broadband connection. The reason for this is that if for some reason the traditional phone network goes down, there is no reason they should not still be able to receive 911 calls from VoIP users as long as the Internet is still functional.
3) Cisco, or some other company in the VoIP hardware business, should work with 911 centers to design a device that will present VoIP calls to the 911 center on trunks that work as closely as possible like normal 911 trunks from the phone company. In other words, a modified ATA-186 perhaps, but redesigned to present multiple lines in whatever configuration is acceptable to 911 centers (note that the R&D needed to produce this could also, with small modifications, be sold to businesses as multi-line versions of the ATA-186). Perhaps such a device even exists already - if so, it needs to be found and identified, and then 911 centers need to install it.
4) It should be made clear that until the technology specified in #3 is available AND 911 centers choose to install it, it is impossible to reliably deliver 911 calls from VoIP users to 911 centers in the way they want to receive them. You might as well ask water to run uphill. But, this is only half the problem.
5) VoIP is inherently location nonspecific. Furthermore, a certain segment of VoIP users actually prefer it that way - they would not want an AT-device capable of "tattling" their exact location in a non-emergency situation. On the other hand, 911 centers AND certain agencies with three-letter acronyms would love it if exactly that capability were present. So the first question to ask it, can we implement any kind of location sensing that would satisfy 911 and perhaps CALEA requirements without unduly compromising customer privacy?
6) If it is decided that some sort of location sensing is desirable, there are a couple ways I can think of to implement it. Each would require serious modifications to the ATA type device, although the first MIGHT be able to be implemented in firmware, if there is room for the code.
7) The first proposed method requires the customer to originally specify the location he is at. In this sense it would be much like what Vonage does today, where they take the customer address and attempt to figure out which 911 center it should reach. It would be helpful in this regard if 911 centers across the nation would join in providing a web site where an address could be typed in, and the database would return all the necessary data associated with that address, that would need to be delivered on a 911 call. If an unknown address was typed in, all 911 centers serving that zip code would be queried (in real time, or near real time) to find out if the address is in their area, and if so, what data should be associated with it. Note that such a database would also be of great assistance to the 911 centers themselves, when they receive a call reporting an event that is out of their area (for example, "Hello, 911? I was just chatting on the Internet with my friend in California and he said it felt like his chest was exploding and now he isn't responding at all, and his telephone just rings and rings with no answer..." With access to such a database, and a correct address, the 911 center would know within a few seconds where to forward this call to.)
In this proposal, once the address is determined, the ATA-186 or similar device would initiate a standard traceroute to a known, highly reliable IP address. It would look at the first non-local addresses (generally the cable modem's address assigned by the ISP) and store that. If that address ever changed, particularly to an address where any of the first three octets were different, it would notify the VoIP provider that the customer had changed location and that the 911 data was no longer valid. Perhaps at that point the ATA-186/whatever could present a very non-standard dial tone to alert the customer that 911 service is not available until he goes to his provider's configuration page and enters the data for the address he's currently at (or indicates that he is either outside the United States, or is at a temporary location where he has other was to contact emergency assistance for the duration of his stay). The provider would keep a list of IP addresses for which the customer has provided address information, so that when the customer moves from one location to another, they would know if that IP were already associated with a known location and would make the change automatically.
8) The second proposal would be to implement some form of GPS in every ATA type device. However there are several drawbacks to this: It's expensive, it won't work in places shielded from radio signals, and it relies on systems not under the control of the VoIP provider. For example, to get reliable GPS information, an ATA-186 would have to either query nearby cell phone towers, something an ATA-186 is simply not designed to do and that would bring howls of protest from the cell providers themselves, OR it would have to have a clear view of GPS satellites, which is probably NOT the case with the majority of ATA devices.
In some cases it might be possible to design a system such that a remote GPS receiver would be mounted on a customer's roof, or other nearby location with a clear view of the GPS satellites, and then transmit the GPS data to nearby ATA devices using a wired or wireless connection, but most users would have neither the technical ability nor the inclination to install such a device. Furthermore, most customers would have no idea what to do if such a device stopped working. For example, a customer might mount such a unit in what seems a great location during the winter months, then be totally clueless as to why it has stopped working in the spring, forgetting that tree leaves obscure satellite signals.
Then there is the issue of how to get power to the remote unit. In any case, including GPS capability would severely drive up the cost of an ATA unit, and reliability of the location information might not be any better than with another system tied to exploring nearby IP addresses. Further, if a customer changes location (especially on a temporary basis), they are not likely to want to climb up on their roof to get the remote GPS sensor.
9) So, given the current technological limitations, 911 centers need to understand that even when address information is provided on a call from a VoIP customer, they need to attempt to verify that the information given is correct, and they need to understand that sometimes a customer may be at a location where they will be unable to help. For example, if a customer takes his ATA unit to a hotel in Siberia and the hotel catches fire and the customer dials 911, it's not likely the 911 operator back home will be able to do much other than offer general hints on surviving a hotel fire until help arrives. Even that small amount of help could make the difference between life and death, though.
The above are, as I said, just a few random thoughts in no particular order. If anything is helpful, feel free to use it. If anything is clear as mud, don't be afraid to ask for clarification! | |
wigwam
join:2001-12-12
Minneapolis, MN
| 911 is a great service very nice. But, who is going to pay for all this integration for VOIP? I can't even remember the last time I or anyone in my household called 911.
For the sake of keeping VOIP on the cheap I can do with out the 911 that a land line would provide.
Until VOIP becomes a standard like pots lines then I would suggest those that have a dependence or require 911 access not venture into the VOIP. | |
usa2k
Please PRAY for Rebekah
Premium,MVM
join:2003-01-26
Canton, MI
clubs:
 ·VOIPo
 ·WOW Internet and C..
·Broadvox Direct
| reply to jeffpulver
I was searching "essential service" in further thought about VoIP 911 ramifications, and found this article interesting: »afr.com/specialreports/report2/2···5ZC.html
This deals more with cellular service, and traditional phones.
Can 911 be taken seriously on anything but a POTS phone? The infrastructure that provides VoIP have more weak links, due to the variability of the ISP and their lack of vested interest in its success.
Although cell phones may not at the moment deal with 911 as POTS does, maybe there should be a basic cell phone in all devices for E911? This device, with no selling features, could be mass produced for that one purpose and go along for the ride with what ever is being sold. (VoIP device, POTS phone, Marine Radio, FRS, ...) Maybe that is too extreme a concept?
If you can buy a microwave oven for $50.00, I would think something like this could be made very cheap, considering that the numbers of these devices could be in the millions and then some!
(Thats my 2 cents!)
--
jim,
Vonage User, iCH user, WideOpenWest, and a fan of LINUX and Windows in that order.
Link to In-Laws gospel quartet! | |
clecrupt9
join:2002-01-22
GA
| I think thats an Aussie article. I guess that if you were in Australia you could make the case that you are not a fixed line service since you could take the ATA anywhere.
Something to chew on...what is the world gonna look like when people here can have Australian numbers and Australians have US numbers and there is full international worldwide voip. How the heck can they expect to have 911 work then? This could be really tough to sort out, were just seeing the start. | |
usa2k
Please PRAY for Rebekah
Premium,MVM
join:2003-01-26
Canton, MI
clubs:
·VOIPo
·WOW Internet and C..
·Broadvox Direct
| said by clecrupt9  :
I think thats an Aussie article...
:) Yes, I guess I went FAR to state my thought  I think the issues are still comparable.
I'm just wondering if ISPs should be considered essential services, and require back-up generators, UPSs, and redundancy that can survive blackouts and other reliability concerns? And then what? VoIP just benefits from this at the ISP's expense?
--
jim,
Vonage User, iCH user, WideOpenWest, and a fan of LINUX and Windows in that order.
Link to In-Laws gospel quartet! | |
WhyADuck
Premium
join:2003-03-05
| reply to clecrupt9
I am convinced that one key to making this work correctly is to require all 911 centers to have a broadband connection and some standardized way to receive VoIP calls. The standards for this may have to be worked out and it may take a while, but the one thing we cannot let the 911 centers do is dictate that all calls must come in over the PSTN. If their attitude is going to be that they won't recognize VoIP, then VoIP should not recognize them. And the reason is simple: Although VoIP providers have little control over what happens on the Internet, at least the Internet is supposed to be robust enough to route around points of failure. But one fiber cable cut can put a PSTN exchange out of service for hours, maybe days in extreme cases. So just as PSTN callers should not have to depend on VoIP to connect to 911, neither should VoIP users have to depend on the PSTN for that connection.
And another reason is that if every 911 center has equipment capable of receiving 911 calls directly off the Internet, then you eliminate all the problems with "My switch is in New Jersey, so how can I possibly connect to a dedicated 911 trunk in Idaho?" Dedicated 911 trunks are intended for local connections, to separate 911 calls from normal trunks which might be congested with normal calls (this sometimes backfires if everyone tries to call 911 at once). But no way are you going to have a dedicated trunk to every 911 center in the country from each VoIP provider's switch.
I'm not saying 911 isn't important, but this is an issue where not only will the 911 centers have to be willing to be flexible, but the equipment makers (and I would say probably Cisco at this point) are going to have to work with the industry to build a device that can go into a 911 center and simulate a group of dedicated E911 trunks.
Just as a shot in the dark, I would say that what is really needed is a device like this:
- Accepts one, two, or three incoming broadband connections and can fully utilize the good circuits if one or two are not connected or down,
- Acts as a normal firewall/router, but also has a specific method for incoming call data from VoIP callers to be sent to 911 center computer terminals,
- And has a number of ports that emulate E911 trunks for incoming calls, with full capability to transfer such calls to a different 911 center should they inadvertently arrive at the wrong center.
- As a bonus they may want to include some "virtual ATA-186's" so the centers can subscribe to a VoIP service for their business-related incoming and outgoing calls. As I mentioned before, this facility could also be reused in a business-class VoIP unit.
Now I'm not saying this is something that could be drafted on the back of an envelope in 30 minutes. You'd have to find out how existing 911 center equipment operates, and then try to emulate the type of trunks it expects to see as closely as possible (or maybe design something better, that would form a complete system and would accept the older PSTN E911 trunks). This isn't going to happen overnight, and that's what states like Minnesota need to understand. | |
jester121
Premium
join:2003-08-09
Lake Zurich, IL
 ·surpasshosting
·ViaTalk
| We're in the midst of installing an 80-station VOIP solution right now, and also wrestling with some of these same issues. The government seems to muddy the waters at every turn as well.
Not trying to hijack the discussion here, but there are also more complex things thrown into the mix here, when you get business service involved. For example, a newer requirement of E-911 is that in larger buildings (I think 40,000 sq ft?) and "campus" environments, specific location data must be sent to the 911 center. So, if I'm in a 100k sq ft warehouse and someone cuts their hand off on the loading dock, the 911 call has to indicate that the calling station is in the southwest corner of the building at 1234 Whatever Road. And at our golf course, a call to 911 has to indicate if it's coming from the pro shop, the driving range building, or the cart barn, since that qualifies as a campus.
Our provider (Verizon/Go-Beam) handles this by providing the 911 center with location data based on the Cisco phone's IP address, and we're responsible for updating them when/if we move a phone from one location to another. Redundancy is handled by an ATA hooked to a POTS line at each site, or in case of a campus environment, a POTS line hooked to an emergency phone on the wall.
This requirement exists up front for basically any phone system. In our case, Verizon handles it as part of the monthly charge. For other solutions, there's a "paperwork" fee charged by Ameritech/SBC -- about $4000 up front for our size, plus $50 a month to maintain the E-911 data. | |
dipswich
Premium
join:2003-06-27
Raleigh, NC
 ·AT&T U-Verse
·VoicePulse for Bus..
·ViaTalk
·RoadRunner Cable
| reply to WhyADuck
To add my random thoughts--
911 works with the traditional telephone companies because the identification is the phone number, and the physical location where the number terminates is known and available to the 911 center. Would the current 911 system work if the phone were mobile? What happens when you call 911 on a cordless phone from the neighbor's house (you know, the one 2 doors down and one street over)? Hopefully, when the responders show up at the handset base station, they'll see the smoke or hear a yell from 300 yards away.
Cellular 911 has bigger problems. At least they (should) know the cell you called from, but quite often the 911 call gets routed to the wrong call center and caught up in interagency problems. Until triangulation is perfected or GPS is in the handsets, 911 from a cell phone is at best risky. How long have the cell companies delayed in getting a viable 911 system implemented?
For VOIP, the IP address of the caller is the identifier. For some IP addresses, reverse DNS lookup could pinpoint the call to a cubicle or address. For many others, the city is the best that the IP can be tracked (maybe the call will at least route to the appropriate call center). Then, there are those records where the IP addresses are only locatable to within a few thousand miles-- corporate headquarters or the ISP's business address, even after whois and traceroute. 911 has no chance! It would be easy if every IP address had a (correct) LOC or GPOS record in DNS pinpointing location. But, I'm not betting my life on that.
So, what is the responsibility of the VOIP provider? Is providing the IP address enough? To whom? A central agency, or does each company mine its own data? Should there be liability if there's ever a network port without a geographical identifier? I don't think it's the VOIP provider's responsibility to answer these questions. Neither is it Minnesota's!
For now, all we can do is admit the problem and ask for the same leeway those wireless companies are exploiting. | |
clecrupt9
join:2002-01-22
GA
| Another thought, it could be entirely possible to have the CPE which so far has been the ATA186 have a user defined register for zip code. You the user place the ZIP code where you physically are and that is relayed to the provider who then knows where you are. Dont trust the Government- then dont put anything in. Now the question becomes how reliable must the provider be when you call 911. Obviously the internet is going to gow down from time to time, acts of God will happen as well that prevent the provider from access 911 ect. Perhaps a blend of having the user program his or her location in, the provider giving its customers access to all "known" 911 response centers nationwide, and if there ever is a VoIP phone book, it too should contain a nationwide listing of 911 and taylor it just for VoIP. In this manner the VoIP company has led the horse to water, making him drink well thats up to the horse. | |
BillRoland
Premium
join:2001-01-21
Ocala, FL
clubs:
·Cox HSI
| I don't know a whole lot on the subject, but it seems to me that the only way to reliably get address information on a VoIP call would be to somehow have the ISP send contact info from the cable/DSL modem. However, I have no idea how this would work.
--
"Don't steal. The government hates competition." | |
clecrupt9
join:2002-01-22
GA
| reply to jeffpulver
Here is another one, how about a 911 IVR that handles SIP, 323, ect. You call 911 and the service provider hands the call off to an IVR where you state your address. Opon stating the address the call is routed to the center in its database. In this way the IVR has PSTN handoff, no need to update the center and no need for the provider to have special 911 trunking. The service provider is responsible for getting you to the IVR only. This takes pressure off the provider and any service provider or business can have access to the database for either a per call charge or a flat fee. I wonder if the broadband providers would be interested in this? Or would they want their own solution? | |
jester121
Premium
join:2003-08-09
Lake Zurich, IL
·surpasshosting
·ViaTalk
| reply to dipswich
said by dipswich :
So, what is the responsibility of the VOIP provider? Is providing the IP address enough? To whom? A central agency, or does each company mine its own data? Should there be liability if there's ever a network port without a geographical identifier? I don't think it's the VOIP provider's responsibility to answer these questions. Neither is it Minnesota's!
That's really the crux of it. If I take it upon myself to move IP phones around between sites and mung Verizon's database, it really can't be their fault -- any more it's Ameritech's fault if I go down to my 66 block and rewire a bunch of stuff they installed there.
I think eventually the VOIP providers will figure out that they're best off providing the service in one form or another, and they'll start factoring the headaches into the price. | |
clecrupt9
join:2002-01-22
GA
| But see thats whats great about an IVR, it doesnt matter what youve done, simply say your address and it will pull the required info based on that. And the service provider can send calls to the IVR via SIP or whatever, the PSTN costs are only at the IVR and the fact that the IVR could pull a local DID within the 911 center's area means that the PSAP receives the call just like any other 911 call. No back line or any other mess. Too bad I dont have the money, Id do this on a trial here in Ga. | |
clecrupt9
join:2002-01-22
GA
| reply to jeffpulver
Some people may find this site informative on 911. I still dont know what would happen if 911 wasnt available or if a carrier was at fault for not properly routing a 911 call. I do know that there are certain things you cant sue the government or a telecom company for. I have a suspicion 911 is one of them. However since broadband phone providers are not "telecom" companies, they probably could be sued for not properly routing a 911 call when they advertise 911 service. I am not a lawyer though.
»www.nena.org/ | |
WhyADuck
Premium
join:2003-03-05
| reply to jeffpulver
Here's a further thought. When Minnesota or some other jurisdiction asks when reliable 911 location information can be provided, one possible response might be "As soon as IPv6 is fully deployed." The problem right now is that there aren't enough IP's available to tie an IP address to a physical location - what appears to be a single IP address on the Internet may very likely serve multiple physical locations (perhaps all within the same building, perhaps spread over a vastly wider area).
When IPv6 is rolled out, it will once again be possible to "waste" IP address by giving each node on a network its own IP address. And this will make geographic location via IP address more reliable by a huge factor.
See, the one thing regulators really can't argue against are technical arguments. How do you think telephone companies pull the wool over the eyes of state PUC's? The give them all sorts of information about costs depreciation, technical capabilities, etc. etc in ways that make the information seem to say whatever the company wants it to say. Most Commissioners are political appointees and unless some opposing group is smart enough to detect the BS and point it out to the good Commissioners in a way they can understand, they will probably just roll their eyes and pretty much let the phone companies have their way.
Well, the thing about IPv6 is fact - it will be much easier to detect when an ATA has changed physical locations once every node, router, etc. has its own public IP address. Not 192.168.whatever.whatever, but an address that stays constant on the 'net. Once that is the case, the ATA device will be able to do a traceroute two or three hops away and determine with fairly reasonable certainty that the user has changed locations. But you can't do that very well with IPv4.
Just another possible way to approach the problem. | |
wigwam
join:2001-12-12
Minneapolis, MN
| What about companies that use proxy servers? and firewalls? The IP address thing nice idea but, sounds like it would need a fully dedicated organization to keep track of all the adds/moves/changes.
I still think the best way to accomplish all of this is to implant GPS and cell phone technology in people. | |
|
