MIT Preps Wireless Network For Cambridge
MIT researchers and volunteers are rolling out a municipal wireless network that will at least initially, be free for low-income residents.
February 8, 2006
A group of researchers and volunteers from the Massachusetts Institute of Technology is rolling out a university-developed broadband wireless network that is planned to cover the city of Cambridge, Mass., where MIT is located.
Unofficially called the Cambridge Public Internet, the mesh network is robust, easy-to-install, and inexpensive for end users, who won't pay to use the network, at least initially.
The effort, which has the backing of Cambridge city officials, is already up and running in a broad section of the city near MIT. "There are 12 users on it right now, as we speak," said Bob Keyes, Wednesday morning in an interview. Keyes is a volunteer researcher working on the project. "For users, this is unlike other municipal networks where you have to put money up."
If the effort meets its goal of deployment by late summer, it will set some sort of speed record for installment of a municipal Wi-Fi network.
Based on an academic research project at MIT called Roofnet, the project has been in test mode around the university's campus since 2003. It has grown more-or-less organically since then in the city of Cambridge. Variations of Rooftnet have been deployed in other cities including Portland, Oregon, and Berlin, Germany.The effort stands in stark contrast to Cambridge's neighbor across the Charles River, Boston, which has resisted any citywide Wi-Fi deployment. There are, however, several Wi-Fi hotspots in Boston, including one based on Roofnet located in a low-income section of that city.
The Cambridge municipal effort had its origins in a desire by Cambridge city officials to provide inexpensive or free Internet access to low-income residents. According to published reports, City Councilwoman Henrietta Davis spearheaded the effort when the local cable broadband provider showed no interest in lowering its prices for low-income residents. In addition, Cambridge's Chief Information Officer Mary Hart got behind the project.Roofnet still has overtones of an academic research project laced with community barn building in the form of assorted academics, researchers, and hackers donating their time to make it a success. Roofnet's core development came out of MIT's Artificial Intelligence Lab and the university's vice president for Information Services and Technology Jerrold Grochow pitched in with valuable support. Another key player is Kurt Keville, a research specialist at MIT's Institute for Soldier Nanotechnologies. Volunteer Bob Keyes is also CEO of networking company XA Net.
'We want to take Roofnet out of the academic world and move it into the community," said Keyes. "We want to do a technology transfer, although officially it's still a test network."
Keyes, who links with Roofnet from an antenna about one mile distant, said the network's mesh technology has been working well so far. The core service is based around a $15 Netgear WGT634U router; the chip inside the router is replaced with a custom-designed chip as well as custom software, much of which is in the public domain. Users with Wi-Fi 802.11b capability on their computers simply log on to the service once and they are up and running, Keyes said.
"Nobody knows for sure how many nodes are up," said Keyes, adding that the mesh network can handle many users. "Theoretically we could have a million users." An important part of the philosophy behind Roofnet is easy access and inexpensive accessibility.The Cambridge plan calls for Roofnet to be free for low-income residents, although there could be a charge later if usage overloads the network. Keyes added that traditional networking equipment suppliers could provide gear for the network at a later date. And, while the origins of Roofnet are in MIT, there have been discussions with Harvard University researchers about joining the project.
What's the most important piece of the Cambridge Roofnet project, Keyes was asked?
"Cost," he replied. "That's the whole point of it."
Add it all up -- very low-cost, robust throughput, quick and easy installation -- and it should be successful, Keyes said.
You May Also Like