More Than 45 Years Old, The 911 Network In The U.S. Needs An Overhaul For The Mobile And Internet Age. Fortunately, A Plan Is In The Works.

(Note: This is an excerpt from Avaya’s recently-published e-book, The 2013 Guide To Collaboration Trends. Download the full 160-page PDF here.)

The very first telephone call was, arguably, an emergency call.  “Mr. Watson—Come here! I want you,” Alexander Graham Bell said into the transmitter on March 10, 1876, after spilling battery acid on himself—or so the story goes. Then there was a 92-year gap until February 16, 1968, when the Alabama Telephone Company processed the first 911 call, beating out AT&T. A week later, Nome, Alaska, deployed a similar network. And thus the evolution of emergency communications as we know it today began in the U.S.

Prior to AT&T’s choice of 911 as a universal emergency number, there were separate numbers for the police, fire department, and ambulance services. Compounding the difficulty for citizens, these numbers also varied by city or state. Besides being a single nationwide number, what made 911 so special was its selective routing design. This provided the ability to route specific telephone lines to specific 911 centers based on the caller’s address, as provisioned at the phone company central office.

For its first 12 years, 911 networks were slowly deployed across the U.S. In the early ’80s, automatic number identification (ANI) was added to 911. Similar to caller ID, ANI identified the caller’s telephone number to operators at a 911 call center, initially on a separate screen. This provided information needed to call back the caller in case the line was disconnected.

The next step was the addition of location information. This was accomplished by the 911 center using the caller’s ANI to query the address stored in the carrier’s database of subscribers. This address, typically the caller’s billing address, was then shown on the dispatcher’s computer screen.

Going Mobile
As 911 systems became more complex, the addresses on file became correlated to map coordinates as well as additional data sources, such as the geospatial XY coordinates that are available from cellular network carriers. This was critical. With the arrival of cellphones in the late ’80s, phone numbers stopped being specific locations on the planet, and began to represent individuals on the go. It was at this point that the logic used to locate callers in a 911 network started to crumble.

To correct the problem, and remain within the capabilities of the existing network, pseudo-ANI numbers (p-ANI) were introduced as “shell records” in the ANI database. These records initially represented the location of the cellular tower to which the caller was connected. Phase II built on this architecture by pulling location data from the cellular network.

When a mobile phone dialed 911, a pseudo-ANI was allocated to that call event, and while the caller was talking to the dispatcher, the mobile network would use a combination of radio triangulation and GPS coordinates to establish a location for the mobile phone. The X and Y coordinates would then be stored in the pseudo-ANI record. Public safety dispatchers would then be able to query the carrier and the cellular location information associated with the call would be provided.

Virtually Speaking
Mobile technology had decidedly crept into our consumer lives. But inside businesses, most users were still hardwired to their desks inside large office buildings. Voice over IP (VoIP) and Wi-Fi were still just fantasies. Once companies started using these technologies in the late ’90s, they also faced the location tracking problem and had to start figuring out how to enable their employees’ locations to be discoverable and identifiable to 911 networks and PSAPs (Public Safety Access Points, i.e., 911 call centers).


Today’s E911 network handles more than 240 million calls per year. But its archaic architecture of routing callers based on telephone numbers gets outmoded the more that technology advances.

This geolocation problem became compounded as virtual private networks (VPNs) encompassing both data and voice became popular and large numbers of employees started working from home. An employee at home in Nebraska logging into her corporate data center in Seattle who calls 911 from her work phone might show up as a Seattle–not Nebraska–caller and thus be automatically routed to a dispatcher in Washington–or not be connected at all.

Today’s E911 network handles more than 240 million calls per year, according to the National Emergency Number Association (NENA). So while it’s effective, its archaic architecture of routing callers based on telephone numbers gets outmoded the more that technology advances in this mobile, virtual age.


The FCC, Department of Homeland Security, and the Executive Office of the President have all committed to building the next generation of 911. But it will take both time and funding to deliver on this vision.

Fortunately, stopgap solutions emerged. The first were Voice over IP positioning centers (VPCs). These networks are able to collect and store employee location data uploaded to them by corporate networks, which they can then provide to 911 call centers across the United States and Canada on demand.

Each VPC is 100 percent dependent upon the enterprise to provide it with accurate information. Fortunately, the intelligent networks deployed in most enterprise offices today do a good job of tracking the location of employees and their devices. They accomplish this by segmenting users into three different groups–heavy travelers, remote teleworkers, and in-office employees. Each of these groups is tracked by segmenting IP addresses, tracking virtual LANs, and other industry standard discovery techniques. All of that real-time data is correlated by the enterprise and then uploaded to the VPC as location data to be presented to the PSAP.

Even better would be if enterprises published real-time location data on their employees to a DMZ (i.e., semi-open) portion of their networks from which agencies or public safety data aggregators such as Smart911 could then download and deliver data to the PSAP. A rising number of 911 call centers are adding data from Smart911, which holds the safety profiles (health records, allergies, drug risks) of a fast-increasing number of consumers. That way, emergency responders can have all of this useful data when responding to an emergency.

Next-Generation 911 

So what’s lacking? Primarily, that E911 remains an analog, voice-only- based network. That means all of the data that we’d like emergency dispatchers to have available must either be transmitted separately over the Internet, which has its own issues (latency, security, etc.), or downloaded from data sources that are often not accurate or up-to-date.

The ultimate fix for this is a new, next-generation 911 network. Its foundation is a modern, secure IP-based network called the Emergency Services IP network (ESInet). This would be an intelligent highway for data to flow directly between all relevant parties: callers, carriers, emergency dispatchers, etc.

By essentially creating a public safety-specific, redundant private network unifying voice and data, the ESInet will cut out the middleman (Internet) and provide a direct, more reliable route for this key 911 data to get to emergency service providers at the same time callers are on the phone.

After years of debate, the professional association most closely associated with 911, NENA, released in June 2011 a document making the ESInet the foundation of its proposed next-generation i3 network. All of the major parties are building to that standard today and collaborating on tests. Europe has also ratified the proposal around its next-generation emergency network, called NG112 LTD (Long Term Definition), which uses a technical foundation very similar to the NENA i3 standard as well as an ESInet framework.

The FCC, Department of Homeland Security (DHS), and the Executive Office of the President (EOP) have all committed to funding and building the ESInet in the U.S. But it will take both time and additional funding to deliver this end-state vision. Yet, the problem exists, and needs a solution, today.