By Sarah Bahari
Next to stacks of papers and old science journals, a sleek, black video phone rests in the corner of Ram Dantu's office. Like a growing number of phones, this one connects callers through the Internet rather than by traditional wire lines.
"In the next five years, the way we communicate will change greatly," says Dantu, an associate professor of computer science and engineering at the University of North Texas. "This is the phone of the future."
That could be a problem for emergency services like 911, which have had trouble keeping pace with the increasingly widespread technology. Reports of delayed responses, misdirected ambulances and unanswered calls from Internet-based phones to 911 are on the rise.
And, like e-mail, Internet-based phones are vulnerable to spam, which could lead to waves of unwanted phone calls and potentially dangerous service tie-ups that could prevent callers from even getting through to emergency help.
With grants totaling more than $3 million from the National Science Foundation, Dantu is leading an effort to overhaul the 911 infrastructure to better handle emerging technology, such as Internet-based phones, text messages and phones with photo capabilities. He also is developing software to detect and block spam from overloading the emergency service system and developing technologies to better alert the public to emergency situations.
The research is timely. Driven by lower costs, more than 24 million people in the United States have switched to Voice-over Internet Protocol — or VoIP as it is commonly called — making adequate 911 services a paramount concern. Within the next decade, Dantu says, VoIP is likely to replace the traditional telephone system.
One of the biggest issues for 911 is how to track phone calls made from Internet-based phones. VoIP is tied to an Internet address rather than a physical location. For example, Dantu says, an Internet user who lives in Denton may call 911 from his or her laptop, but that person could physically be at a park in Dallas. Tracking that person is a challenge.
That was illustrated last year when an 18-month-old boy in Canada died after his parents used VoIP to call 911 and report he was having seizures. The call was routed to the 911 center near the family's previous address about 2,500 miles away from their current home.
"This is the central challenge, and it affects all of us," Dantu says. "Delays in emergency response times are a life-and-death issue."
Researchers from UNT, Columbia University and Texas A&M University are collaborating to develop a testbed that will enable further research and analysis of the next generation of 911 services.
A testbed is a platform for experimentation used in large-scale development projects that allows for rigorous testing of scientific theories, computational tools and other new technologies. The term typically refers to the integration of several network elements, including hardware devices and software programs.
In this case, the testbed will help researchers develop a strategy for tracking phone calls made from Internet-based phones. UNT is leading the project as the three universities divide duties to build the testbed.
Dantu is creating a 911 testbed for future research to experiment on the next generation of 911 services. With the most recent grant — $1.3 million awarded in summer 2008 — he is leading a team of researchers from Columbia University and Texas A&M University to design and build a platform.
Dantu and his team of researchers are searching for a low-cost and practical way of tracking callers. Possibilities could include location-sensor programs, similar to global positioning system (GPS) technology or cellular-tower tracking; networks capable of acquiring and storing location information; and delivery of location information to Ethernet or wireless connections.
Another possible approach is to use Bluetooth-equipped devices that can deposit location information within a building and transmit it to mobile users.
Because each of the location-determining systems has positives and negatives, the best option will likely be a combination of several sensor technologies, Dantu says. For example, GPS works well outside but is not reliable indoors. Finding a way to deliver locations to Ethernet connections could resolve the indoor problem, but doing so would require system administrators to map every Ethernet jack and wireless access point, an arduous task.
Dantu and fellow researchers will use computers to create software that can support and integrate multiple location-sensor systems. They will then compare the systems, measuring accuracy, reliability, cost and the energy required for determining location information.
Dantu also is studying other problems presented by VoIP. How can 911 call centers be secured from outside attacks that would tie up available lines? What can be done to ensure service during large-scale emergencies? How can neighborhoods be quickly notified of emergencies? And how can 911 services for people who are deaf or hearing-impaired be enhanced through the use of video phones?
Karl Levitt, program manager for trustworthy computing at the NSF, says Dantu's quick and pragmatic approach to VoIP threats has made him the national leader in this field.
"Most people, particularly in industry, look backward to deal with the threats we experience today," Levitt says. "What I like about Dr. Dantu's work is he's looking ahead to future security threats."
Dantu's research recently caught the attention of the Metroplex Technology Business Council, the largest technology trade association in Texas, which named him one of four finalists for its 2008 Tech Titan Award in the innovator category. The award goes to a person or group of people responsible for breakthrough ideas, processes or new products.
In addition to research on security issues and research that will lead to a better 911 system, Dantu's UNT lab is working on several phone features that could soon be as basic — and industry changing — as caller ID and call waiting. Possibilities include cell phones that regularly report the user's location to friends and family, or phones that will predict with 90 percent accuracy who will call, long before they do.
Santi Phithakkitnukoon, a doctoral student who works with Dantu, is developing software that predicts incoming and outgoing phone calls. For incoming calls, the software will study phone habits, then generate daily lists of predicted calls. For outgoing calls, any time the user attempts to make a call, a list of the most likely contacts will pop up to reduce the searching time.
Early this year, Dantu's lab installed the software in phones of about 100 UNT students, who will try it out for a few months. Dantu and fellow researchers then will study results and make improvements.
"This will help people schedule their lives much better," Phithakkitnukoon says. "They can prepare for work-related calls. They can know when their family and friends might call. They will be able to look at their phones and plan their day."
The lab already has developed software that automatically tracks the user's location and sends a message to Facebook, the online social networking web site.
"We basically want our phones to read minds. No one wants a phone ringing off the hook at the wrong time," Dantu says. "We're making the phone as smart as possible."
Judge coordinator Art Roberts says Dantu has shown "a wonderful foresight" in addressing the issues with VoIP.
"Dr. Dantu saw a problem and he took steps to solve it," he says. "He deserves a great deal of credit for getting the ball rolling on these serious issues."
VoIP allows users with a computer and standard Internet connection to make toll-free calls anywhere in the world. Companies like Vonage and AT&T offer the service.
But with VoIP's vulnerability to spam posing a nuisance for homes and offices and real trouble for 911 operators, Dantu's team is researching solutions. Paul Sroufe, a graduate student in computer science and engineering, is developing software with Dantu that could detect and block unwanted phone calls. For example, the network could notice hundreds of phone calls coming from one address at the same time and assume they were not legitimate calls.
The researchers are modeling the system after spam filters commonly used for e-mail.
"Imagine one telemarketer making unwanted phone calls, and what a nuisance that is," Sroufe says. "Now imagine 50,000 computers automatically sending spam. We must find a way to block those calls, while making sure not to prevent legitimate calls."
Phone calls won't be the only evolving technology. Emergency centers also must prepare to receive text messages and photos. Eventually, someone could witness a car accident and send a text message with an attached photo, rather than just calling, Dantu says. That would help 911 operators know how many ambulances to send and what sort of medical assistance is needed.
Dantu also is working on technology that would provide reverse notifications. In an emergency, 911 could send phone or text messages to residents of certain neighborhoods.
"From developing new technology to ensuring 911 can find us in emergencies, these communication issues affect all of us," Dantu says. "The way we communicate will continue to evolve, and we must be prepared for these changes."
The Future of Lighting
Researchers look to organic light-emitting diodes for energy-efficient, money-saving light.
By Sarah Bahari
Next Generation Learning
UNT leads transformation of large classes with new technology and active learning.
By Alyssa Aber
UNT-Chile Field Station serves as nature's laboratory for biocultural conservation.
By Randena Hulstrand
Solutions By Design
Braille readers, Haitian farmers benefit from UNT communication design.
By Ellen Rossetti
Preserving Endangered Languages
Linguist's electronic archive will help a minority language of India live on.
By Nancy Kolsti
Student examining effects of radiation on stainless steel wins DOE support.
By Mellina Stucky
Emerging strengths, expanding expertise
Research funding, Discovery Park, new deans
Honors for research and creativity
Solutions and scholarship across disciplines
Experts in conducting, Latino political behavior, tinnitus, materials science
Projects in education, science, social science, art, business, music
Poetry, justice, Aztec history and more
A growing research agenda
Web page last updated or revised: February 2, 2009
Questions or comments about this web site? firstname.lastname@example.org
"University of North Texas," "UNT," "Discover the power of ideas" and their associated identity marks, as well as the eagle and talon graphic marks, are official trademarks of the University of North Texas; their use by others is legally restricted. If you have questions about using any of these marks, please contact the Division of University Relations, Communications and Marketing at (940) 565-2108 or e-mail email@example.com.