Traditional 9-1-1 systems, which date back to 1970s, support only voice, while non-emergency communications now feature other media. Adding additional media for 9-1-1 presents opportunities and challenges. Text messages, images captured by cell phones, video clips, and automatic crash notification messages can dramatically enhance the 9-1-1 services by expediting emergency responses and reducing crash clearance times. The rapid increase of residential, nomadic and mobile VoIP usage requires the development of VoIP-based next generation 9-1-1 systems and services that will replace the current circuit-switched 9-1-1 systems. Beyond limitations in media and mobility support, existing systems are inefficient and cannot easily accommodate new functionality. This project is a collaboration among University of North Texas, Columbia University and Texas A&M University, where UNT will be the lead institution, to develop a testbed that will enable research on understanding and analysis of next generation 9-1-1 services. The testbed make possible research and development in reliability, security, function-appropriate privacy and other areas that already difficult in large scale VoIP, but which become daunting when the VoIP system is critical infrastructure.

The broader impacts of this project are many. A testbed for Internet-based 9-1-1 research is particularly important as both state and federal governments are in the process of planning next-generation emergency communication platforms, unfortunately often without adequate vendor-neutral testing and evaluation. Users of the testbed will investigate issues related to locating 9-1-1 callers, securing Public Safety Answering Points, ensuring continuous availability of 9-1-1 services during large-scale emergencies, predicting emergencies, providing citizen alerts ("reverse 9-1-1"), and improving inter-agency coordination. The PIs expect to translate results from research on this infrastructure to engineering guidelines and disseminate results across government organizations, standards bodies such as IETF and National Emergency Number Association (NENA) and 9-1-1 centers. Moreover, the findings from the experiments in this project will be useful for the residents across USA.

Project Report

Current citizen-to-authority emergency systems are based on traditional telephony that uses circuit-switched networks. However, there are new standards, such as the National Emergency Number Association (NENA) i3 architecture and Internet Engineering Task Force (IETF) protocols that support emergency communications over packet switched networks, also known as Next Generation 9-1-1 (NG9-1-1). With NG9-1-1, all communication is based on computers and software systems (e.g., softswitches) that use the Session Initiation Protocol (SIP) as the signaling protocol and Location to Service Translation (LoST) as the call routing protocol. NG9-1-1 calls allow the communication of video, voice, images, text messages and any type of data (e.g., telemetry data). Using an NG9-1-1 testbed, our research focused on testing the LoST protocol for mobile users, implementing a secure signaling framework for SIP and evaluating its performance, transferring calls to first responders, interoperability tests, and training. Testing Location to Service Translation (LoST) for Mobile Users We tested the operation of the LoST protocol when smartphones do the caching of the service boundaries data. This allowed us to first evaluate the impact of the delays of the 3G access network in the LoST queries. To avoid the delays of new LoST queries every time a location changes, we further investigated the specific issue of LoST and service boundary crossings. We presented a new method to determine service boundary crossings, the Minimum Distance method, and compared it with the Ray-Tracing method. Our results showed that the new algorithm reduces the processing time at the mobile LoST client. A Secure Signaling Framework for NG9-1-1 Calls Security is an important concern in the NG9-1-1 architecture. An approach to securing the signaling, such as the Session Initiation Protocol (SIP) and Location to Service Translation (LoST) messages, is to use Transport Layer Security (TLS). Our main contribution is to be the first to implement and test SIP over TLS, IPSec, and secure LoST transactions in an NG9-1-1 testbed. The IETF recommends the use of TLS as a primary mechanism to secure the signaling in IP-based emergency calls. Under normal network conditions our experiments showed that the TLS handshakes in the NG9-1-1 testbed consist of approximately 20 percent of the call setup delay, comparing with calls over TCP. By testing secure SIP and LoST transactions in different scenario combinations, we also observed that better performance was obtained when the client provides routing information (i.e., client does the initial LoST query). Transferring NG9-1-1 calls from the ESInets to the first responders In 2010 Harris County, Texas (which encompasses the City of Houston) installed equipment at the labs at Texas A&M University on the Public Safety Broadband Network (PSBN), also known as Firstnet. Since then, the Internet 2 Evaluation Center (ITEC) lab has been able to obtain additional functional elements, such as an IMS Core, an Automatic Multicast Tunneling (AMT) router which supports multicast video on LTE networks, and a carrier grade SIP proxy. We have documented the signaling scenarios and all the message flows when the PSAP transfers the NG9-1-1 call to the first responder. We assumed the PSAP uses conference/bridging and the SIP Refer method to transfer the call to the IMS domain. Interoperability Tests The ITEC staff supported NENA Industry Collaboration Events (ICE). ICE 1 was held at Texas A&M University, ICE 2 and ICE 4 were held at the AT&T training facility in Dallas, Texas. Texas A&M provided the technical support for these ICE events. Training Undergraduate education was also a focus on this project. We offered one special topics class on NG9-1-1 for senior undergraduate students, and sponsored two Capstone design projects. The first project was the development of a robotic platform that searches for a target object (e.g., an explosive object) and does an NG9-1-1 call with the robot’s location information, once the target object is detected. The second project was the development of a wristband that continuously monitors the user’s heart rate and has the ability to call 9-1-1 and send the user’s latest heart rate readings to NG9-1-1 call takers.

Agency
National Science Foundation (NSF)
Institute
Division of Computer and Network Systems (CNS)
Application #
0751118
Program Officer
Jeremy Epstein
Project Start
Project End
Budget Start
2008-09-01
Budget End
2013-08-31
Support Year
Fiscal Year
2007
Total Cost
$307,000
Indirect Cost
Name
Texas A&M University Main Campus
Department
Type
DUNS #
City
College Station
State
TX
Country
United States
Zip Code
77843