You’d be describing the world of wireless university campuses—a fast-growing world facing its own unique challenges.The spread of Wi-Fi networks in higher education has been so rapid in recent years (especially since it became a standard laptop feature) that many sources peg penetration at 100 percent. Actually, it’s “only” 77 percent, reports Stephanie Atkinson, analyst at In-Stat/MDR, a market research firm in Scottsdale, Ariz. About 14 percent of universities report having full-campus coverage, while the others are content to cover class buildings, dorms and select public areas.
Of course, one reason for the rush is simply that wireless networking is cheaper than cabling—Atkinson cites an average cost of $75,000 for wiring a university building, versus $9,000 for adding wireless access points. But most wireless campus networks overlay a pre-existing wired network, so there must be more fundamental reasons for the rapid un-wiring.
Dan Updegrove, VP of IT at the University of Texas at Austin, identified the main one. “Most universities are moving aggressively toward supplying more and more online library services, online course resources and online administrative services,” he said. “That being the case, we’d like the students and faculty to have 24/7 access. But if they buy desktops and leave them at home, they can only queue up at public access computers while on campus. So it makes sense to create an environment where they will chose to acquire laptops and bring them to campus. And the strongest incentive is to insure they can connect almost anywhere they sit down.”
And of course, there’s customer demand. “In the graduate business school, every seat in the lecture halls has an Ethernet port and a power plug for the student’s laptop, but when the class of Fall 2003 arrived, not one of them had an [Ethernet] patch cable for their laptop,” recalls Kevin Baradet, chief technology officer for the Johnson Graduate School of Management at Cornell University in Ithaca, N.Y. “The cord was one more thing to carry around, and they did not want to do so. It was wireless or nothing. I did not expect the tipping point to come so quickly.”
Two years ago you rarely heard questions about wireless access from incoming freshmen during orientation, agrees John Bottum, VP for IT at Purdue University in West Lafayette, Ind. Now, it’s part of the standard campus tour.
Hot Spot for Teacher
While classrooms are normally the focus of attention in education, when it comes to wireless coverage, public and residential spaces where studying occurs are where colleges are focusing. Though some teachers build their classes around online material, others see wireless access in the class as a distraction. Having Internet access in classrooms is also problematic during exams. Several users said they used software
from AirWave Wireless in San Mateo, Calif., or BeaconPoints from Chantry Networks of Boston, to regulate class use. Others reported writing their own management software with student labor—or, they simply turned off the classroom’s access point during exams.
But other sources spoke of covering 300-seat auditoriums, requiring multiple access points. Sources agreed that the maximum number of users per access point is either 25 or (more typically) 30. However, Ann Sun, senior manager of wireless mobility marketing at Cisco Systems, suggests 10 to 20 users for data and only six or seven when using VoIP.
“When a vendor says their access points can handle 30 people at a time with no downtime, I say yes, in theory, and plan for 15,” cautions Nasri Mukhar, network specialist at the University of Miami. “It depends on the cards in the laptops, and their connection speeds, and whether the machine is infected. An access point tries to accommodate all speeds, and if you are coming on at 11 megabits and I am coming on at 2 megabits, the access point will go to the lowest speed to accommodate everyone. You can’t force people to get the best card, but some will get a $20 card and expect 100 percent connectivity.”
Others said that it was in vain to attempt detailed planning and that only on-going tinkering could meet demand. “You never know how far the signal will go until you install the access point,” says Updegrove. “If the director of the library wants all public areas covered by wireless, it is not immediately obvious how many access points you will need or where you should put them. You could check out a new installation at midnight, and it would behave differently during the day because people are there and people are basically columns of water and cause the signals to propagate differently. It takes a lot of trial and error.”
Of course, the possibility of saturation was in the back of many planners’ minds. Bottum at Purdue says that with 38,800 students and 1,200 access points, the number of simultaneous users has reached about 5,000. “We anticipate getting swamped, but we bought the kind of access points that we can switch from 802.11a to 802.11g and add bandwidth.” The average user session is about 60 minutes.
Baradet notes that the main problem at Cornell was not bandwidth—users know they can plug into an Ethernet port if they need maximum throughput—but the laptops themselves. “The problems were the people who bought laptops with an early version of the chipset or with the drivers not updated. Addressing those issues made most of the problems go away.”
A few campuses get around the problem by controlling the hardware. For instance, at St. John’s University, a private Catholic school in New York City, every student is issued a laptop. It’s paid for with their tuition dollars over four years, and the students get to keep them upon graduation, explained a spokesman.
Sources often complained about the slow arrival of wireless security protocols, but all used some form of authentication to control access. The University of Texas additionally scans each laptop for viruses as it logs on. Many campuses have a lot of visitors with laptops and are making considerable efforts to provide the visitors with temporary access.
Another problem that may be unique to higher education stems from the fact that access points now cost less than some textbooks. Students and faculty are known to plug a “rogue” access point into the campus network whenever they want to read under an
un-served tree. Campuses typically have management software that lets the staff spot rogues. Updegrove says he finds several
a month. Greg Murphy, head of AirWave Wireless, said he
heard of a 400-student college dorm where the administration installed 25 access points—and found another hundred rogues already operating.
Demand and Supply
Meanwhile, academics may not be the only activity on campus that creates a demand for wireless access. At Purdue, Bottum says he received private funding to install a wireless network at the football stadium and to pass out 350 wireless PDAs during games. The network offered sports information during the game and traffic information for planning the trip home. After three games, the log-ons exceeded the number of PDAs that were loaned out. (Only two loaned PDAs failed to come back, Bottum added.)
Higher-education campuses are not the only ones acquiring wireless networks. For instance, schools of the Paradise Valley Unified School district in Phoenix hand out wireless palmtops in elementary school, handhelds in middle school and laptops in high school, explains Jeff Billings, director of IT for the district. Its 44 campuses are also wired, “But we have found that the functionality of wireless blows wired away,” Billings says. “The kids are not confined to their seats and the teachers are not confined to their desks.”
But the support budget is so tight that he only has 10 technicians for 15,000 devices. “Education is only now learning about total cost of ownership,” Billings adds.
And things may not get much better soon. According to figures cited by the National Education Association in Washington, D.C., spending on technology fell in U.S. schools in 2003, from a per-student average of $118.25 in 2002 to $88.59 in 2003. But the installed base continued to grow, especially in the case of wireless networks. While 15 percent of schools reported having wireless networks in 2002, 27 percent were wireless in 2003. The number with laptops inched from 36 to 43 percent. Those that provide PDAs to students grew from only seven percent to eight percent, and the number that provided them to students grew from three percent to four percent.
But Billings ran across a fact that may make it easier to make a case for wireless. Thanks to their small screens, lack of fans and trickle-current battery chargers, the mobile units consume significantly less power than desktops, and power consumption has fallen 33 to 40 percent, he notes.•
Lamont Wood is a Texas-based writer. He can be reached at [email protected]