Unwiring Big Apple Schools - Page 2
November 19, 2002
While the objectives for the NYCDE wireless project are a little different than in most jurisdictions, the department is still reaping some of the more customary benefits. The only other way to achieve the same level of connectivity would be to install switches and routers in each classroom, but that would have been prohibitively expensive.
"We estimate that a wireless deployment can be done in a school at as little as one third the cost of a hard-wired solution," Eaione says.
"But that's not referring to the classical savings over putting a [wired] drop in every classroom," chief wireless architect Jim Anilowski hastens to point out. "We wouldn't even consider putting wireless in a school that didn't already have a wired Ethernet drop."
The NYCDE's unique approach required a unique network architecture. Each classroom is a self-contained wireless extension to the school's wired network. With many if not most wireless LANs, designers look for maximum coverage from each access point. Here, designers had to find ways to limit coverage to just one room. It wasn't always easy.
"In order to achieve a microcell structure, we initially thought that using a 0-gain antenna with an access point capable of reducing RF output power would have been sufficient," Eaione explains.
The department's choice of access points was partly based on this requirement. The Cisco 350 series units have detachable antennas and configurable RF power.
There were other reasons for choosing the Cisco units as well. They're "plenum rated" (fire retardant and toxic emission rated for use in air vents). They also support Power over Ethernet (PoE) and come with intelligent power switches, eliminating the need for plug-in transformers in the classroom. This usually one source of savings over wired approaches.
The initial microcell design didn't entirely work, however. The problem was coverage "splashing" out of one room and interfering with the networks in adjacent rooms, either beside or above or below -- classrooms are packed close together in many of the schools. Sometimes coverage splashed outside the school, causing security concerns.
"We ultimately had to use 9db attenuators in line with all antennas in an effort to produce an appropriately sized microcell structure," Eaione explains.
The department also abandoned attempts to use 802.11b for outdoor bridges between school buildings. It found that the bridges inevitably interfered with the classroom networks. Now it's investigating using 802.11a which uses 5.2GHz U-NII-band spectrum and thus won't interfere with 2.4GHz 802.11b.
Despite these set-backs, the implementation has gone surprisingly smoothly. This is partly thanks to the careful preparation and design work, Eaione says.
"We spent a considerable amount of time planning every detail of the architecture and implementation. For example, we obtained floor plans for every location and digitized them prior to deployment, giving the surveyors a method to document where each access point was going to be located."
The department is also now experimenting with voice over WLAN at its headquarters at Tweed Hall in Manhattan, though it's not apparent it will take this experiment much further in the near term.
While the NYCDE is cleary at the leading edge of Wi-Fi in K-12 education, it is surprisingly reticent about pedagogical results so far. In response to our questions, we received only this apple pie & motherhood statement:
"We believe the wireless networking solution will add unlimited flexibility and usability compared to previously tethered computers, giving our educators a great new tool to help teach our children more creatively and efficiently."
It may be there are in fact few if any immediate pay-backs. This is a very long-term infrastructure project, which may not deliver a return until somebody decides to give every fourth-grader a wireless laptop.