State of Emergency: VoIP and 911

By JD Rosenberg, Speaking SIP

Voice over IP is a technology that has had to jump numerous technical hurdles on its way towards success. Voice quality, ease of use, and NAT traversal have been three of the biggest so far. The next one, which VoIP is in the midst of handling, is emergency services.

VoIP (define - news - alert) has the ability to far exceed existing wireline and wireless emergency services in terms of quality and features. With VoIP, wideband speech can be utilized, providing extremely high quality sound. This allows emergency call takers to hear background noises, which can provide additional information about what is happening where the caller is calling from — a good thing for a trained ear to hear. It can also be beneficial for separating fraudulent calls from real ones. Video can allow for fire and police to see what’s going on at the site of the emergency before they even get there. Text-based media streams allow the hearing impaired to have improved access to emergency services. IP-based networks supporting VoIP will also allow emergency calls to carry additional information about the caller — links to their medical records, for example, to allow responders get the right help to the scene. VoIP will thus be an enabler that allows us to save more lives.

But that’s the future. Today’s state of emergency services is only part of the way towards this vision of the enhancements that IPbased networks could provide. Many VoIP providers do offer emergency services, but it is not yet ubiquitous. Fixed access is provided, but mobility is not yet supported. If the user does move it around, they have to manually update their location with their service provider. Though calls can be routed to the correct emergency service center, in many cases the caller’s location is not provided to the operator. Though connectivity exists to most emergency service centers, in some cases it is done through the “backdoor” — the administrative lines that are meant to be used for non-emergency connectivity.

What stands in the way of removing these limitations and going beyond what can be done today? There are three factors in play:

1. Properly routing calls, and then connecting them to the right emergency services center, requires access to databases and closed access points in the PSTN that have not always been available to the VoIP provider.
2. Emergency services require the geographic location of the user (caller) to be determined, a capability which is virtually nonexistent in today’s Internet.
3. The Internet allows the separation of access from service, so that one provider can offer the VoIP service, while another provides IP connectivity.

The first issue can be remedied in part by defining new IP-based access to emergency service centers, and providing standardized IPbased access to the databases needed to route calls to those service centers. Fortunately, work is underway to achieve these goals in several standards bodies, including the IETF. The IETF has defined a new protocol, called LoST (Location to Service Translation), which provides a simple XML-based interface that can be used to map a callers location (included in the call setup message) to a SIP URI for reaching the emergency service center for the caller’s location. The IETF has also defined standards for representing location information in SIP messages.

The second issue will require both devices and the network to be upgraded to be “location aware”. The technologies that will play a role in that will be highly variable, including GPS for certain devices, link layer technologies such as LLDP-MED for other devices, and new DHCP options for other types of devices. Unfortunately, providing location isn’t just a technical challenge, it’s a business one, and that challenge is intimately coupled with the third big issue.

Perhaps the greatest strength of the Internet is that allows a complete decoupling of the underlying network from the applications that run on top of it. Consequently, the user can access an IP-based application from anywhere that IP connectivity exists. This means that, for a given phone call, there are really two providers in play — the one that provides the access network, and the one that provides the application service. Unfortunately, proper operation of VoIP requires participation from both of these providers. The provider of the access network is the only one that knows the location of the user, but the VoIP application provider is the one that needs to route the VoIP call based on that location information. For the system to work, the access provider must allow the VoIP provider to obtain and use this location information.

Why is this unfortunate? It is unfortunate because, in many cases, there is little business incentive for an access provider to give this location information to the VoIP provider for the purposes of completing the emergency call. Clearly, for privacy reasons, users wouldn’t want this information to be given out arbitrarily without their consent. However, the user will need the information to be provided for the emergency call to complete. If the access provider also has a VoIP service, they would be helping a potential competitor offer better service. If the access provider doesn’t offer VoIP service, it would require them to deploy infrastructure for applications they aren’t providing. It is possible that VoIP application providers can work out business arrangements to compensate access providers for location information. If that cannot come to pass, government regulation might be required to help move things along. To be successful, those regulations must recognize the separation of roles and responsibilities between the access and application providers.

But, fear not! Just as VoIP has handily dealt with numerous challenging hurdles in the past, this one too will be met and overcome. Once it does, VoIP will not only change the way people communicate, it will save lives in the process.

Jonathan Rosenberg is co-author of the original SIP specification (RFC 3261). He is currently a Cisco Fellow and Director of VoIP Service Provider Architecture for the Broadband Subscriber Applications Business Unit in the Voice Technology Group at Cisco Systems. (quote - news - alert)