Default | NEWS

5/1/2007

TFA Wireless: a working example for Houston's mammoth Wi-Fi project

BY JADE BOYD
Rice News staff

When Houston City Council this month approved a plan to blanket the city with a 600-plus-square-mile wireless network by 2009, the city sprang to the top of the list of cities that have committed to providing ubiquitous broadband Internet access. Houston's network will dwarf the 52-square-mile Wi-Fi network in Taipei, Taiwan, which is today's largest, and Houston's plan calls for a larger network than all but a handful of regional and or national networks that are still years, if not decades from being built.

But as Houston city leaders look for examples of wireless technologies, applications and systems to emulate, they won't be casting their eyes toward high-profile test systems in Philadelphia, San Francisco or Taipei. Instead they'll be looking at Houston's Pecan Park, a working-class eastside neighborhood where Rice University researchers and workers from the non-profit Technology For All (TFA) are riding a two-year wave of phenomenal success in providing wireless broadband.

"No one has anything like this," said Rice electrical engineering professor Ed Knightly. "We've got ultra high-density usage. Our user densities are higher than anything I've heard about anywhere in the world."

In wireless broadband, user density is a big problem, particularly with the kind of technology that's being used in Pecan Park, which is called "mesh" networking. In a mesh network, dozens of interconnected wireless transmitter/receivers called nodes are set up, typically on utility poles, to provide service over a wide area. Users connect to one of the nodes, and when their computers send or receive data from the Internet, the information is passed from node to node until it reaches a central hub that has a wired connection to the Internet. The name "mesh" stems from the fact that data sometimes makes multiple "hops," bouncing through the "mesh" of nodes before it finds it's way to the wired hub.

On a mesh network, all users share whatever bandwidth is available from the wired hub. If the wired connection is large, then in theory there should be no problem. But in practice, Wi-Fi networks using today's off-the-shelf technology have a problem: users that are one or two hops away from the hub tend to hog an enormous amount of bandwidth, leaving only a trickle for users farther away. The more users that log onto the system, the worse the problem gets. Knightly's research group solved this problem by developing transit access points, or TAPS, nodes that combine new hardware and programming rules to better share bandwidth.

"We just celebrated our 2,000th user joining TFA Wireless," Knightly said. "Our user density is more than 650 per square kilometer right now, and the goal is to get that up to 1,000."

Knightly said doctoral student Joseph Camp has installed 18 TFA nodes since 2004 and expects to complete the network with about least eight more over the next year. Once finished, TFA Wireless will cover about four square miles of Pecan Park.

The network was paid for by TFA and by a National Science Foundation program that is developing wireless networking technologies for the future. Knightly said the network has been successful as a testbed for new technology – like those aimed at solving the user-density problem – but there's more to the success than just technology.

"One of the unexpected benefits has been that this has spurred other researchers in Houston to take a greater interest in Pecan Park," said TFA President and CEO Will Reed. "The TFA Wireless project opened the door for research that looks at how people will use this technology and how municipal wireless will impact communities."

One research program, spearheaded by scientists at the Abramson Center for the Future of Health, a collaboration of the Methodist Hospital Research Institute  and the University of Houston College of Technology, involves low-cost, handheld cardiac output sensing units about the size of a Game Boy. Methodist plans to partner with the neighborhood of  Pecan Park to design a study that allows people with congestive heart failure (CHF) to monitor their heart condition at home. The devices, which will connect to the Abramson Center using the wireless network, will automatically let patients know when their heart is working harder to pump blood. This is an indicator of decompensation in CHF patients, which is usually followed by clinical symptoms that often land patients in the emergency room. The research center’s goal is to extend the quality of life of CHF patients by giving them early warning of decline, and therefore, a chance to slow the progress of this chronic disease.

The project includes a study by anthropologists at the Abramson Center.

"When we began our project in May 2006, we were interested in untangling the issues of trust and health care in the Pecan Park community," said Abramson/UH research professor Jerome Crowder, who is directing a four-person team of ethnographers who are conducting long-term interviews and focus groups in the neighborhood.

"Based on our initial results, we found that there is very poor communication between residents, their civic association and the city," said Crowder, who holds joint appointments at the Abramson Center and the Texas Learning and Computation Center. "Now, we are advising Abramson Center as well as TFA as to how residents approach, understand and use technology – not only the wireless Internet created by Knightly's group, but also computers, cell phones and all types of technology."

With Houston's citywide wireless plan now on the fast track, the lessons learned in Pecan Park by researchers from Rice, Methodist and UH couldn't come at a more opportune time.