By Rusty Bruns
Charleston Southern University (CSU) has the largest wireless network in the state of South Carolina, with over 600 users. All residence halls, the library, and a mobile classroom are wireless. By the end of the calendar year, CSU will add two additional buildings to the wireless-enabled list. This accomplishment and the provision of free wireless network cards for students to use without any Internet connection costs were both accomplished with a budget under $200,000.
In spring 2000, CSU's MIS department realized that the student dial-up Internet access that had been in place for two years was too slow for the number of users and was no longer a viable solution. The quotes for wiring the residence halls with Category 5 data lines and ancillary equipment exceeded $250,000, not including the added support needed to maintain the student IT services. In an effort to provide the best service to students and to be cost-conscious, the MIS department decided to experiment with wireless technology.
In summer 2000, CSU installed the first Avaya access point on campus in one of the residence halls. It had 10 users. The cost was negligible at approximately $2,000 for all the equipment necessary to test this new technology. To get connectivity, a Category 5 data line was installed from the access point into the electrical closet hub. In the closet, a fiber optic cable carried the signal to the demarcation point from the hub to the router. For the next several months, the 10 trial students reported to MIS regarding connectivity, uptime, and speed. Their reports were very positive and the demand from the rest of the student body continued to mount as more students heard about the trial.
In spring 2001, the school's cabinet approved a plan to equip all the residence halls with wireless technology and hire a full-time technician to maintain the system and support students. The estimate for the project was an $180,000 one-time cost. Monthly recurring costs of $3,500 covered the T-1 managed line and paid for the full-time employee's salary.
This plan also included the cost of purchasing wireless cards for students. MIS purchased only PCMCIA (laptop) or USB (desktop) wireless adapters. To accommodate desktop computers absent of USB ports, MIS set up a partnership with a local vendor to install them. The cost was under $40, which included installation and the USB card.
To use the wireless system, students receive wireless network adapters after signing a custody card (a form the student fills out acknowledging that if the card is lost, stolen, or broken, the student is responsible for the replacement cost). In this way, a wireless network adapter pool was created out of PCMCIA (laptop) cards or USB (desktop) devices.
In June 2001, installation started. Thirty access points, a switch, a high-bandwidth managed cable modem, router, server, and associated equipment were fully tested and operational by mid-August. In fall 2001, as students arrived and were issued their cards, the process appeared to be going according to plan. That is, until the ISP-provided router began locking up. The administrator had to reset it every 20 minutes at certain times of the day.
Further investigation revealed that the router furnished by the ISP was made for an office of 25 to 50 users. The MIS department also discovered that the students were not only looking at Web pages but also downloading movies, music, streaming video, and other media. In fact, one student was using 33 percent of the bandwidth on media downloads alone. By the time this problem was discovered, the number of students using the new wireless system had reached 502, with approximately 40 users hogging all the bandwidth.
The ISP was unable to correct the router issue, and the MIS department made a decision in September to switch to a managed T-1 using BellSouth. MIS also decided that a small router wasn't adequate to handle the load of 500-plus students. To correct this problem, MIS made an investment in a Cisco Pix 515 firewall. In addition to these changes, MIS purchased a bandwidth packet shaper, Packeteer, to control the bandwidth usage. Packeteer enabled the administrator to block downloads of movies, music, and streaming video. It also allowed a limit to be placed on the amount of bandwidth each student received. As a result, all the students were able to enjoy high-speed access to the Internet. The speed to access Web pages and downloads now ranged from three to four times that of dial-up with a total of 600 wireless users.
During the acquisition phase of the project, Avaya rendered rebates for equipment purchases over $18,000. These excess funds (coupled with the significant decrease in the cost of an access point) allowed MIS to install wireless equipment in the library. Wireless technology is also being used in CSU's mobile classroom, which has 16 laptops, a laser printer, and an access point integrated into a cart.
A totally wireless campus
Now, a wireless chart is being created to identify the external reach of the internal access points. Initial data shows that the external reach exceeds 100 feet beyond the exterior block walls. An external antenna has been purchased to experiment with the possibility of a totally wireless campus for students. CSU is also testing a Compaq IPAQ with a wireless adapter as an alternative to laptops and another option for students with desktops. The initial results of these tests are promising and the campus may be completely wireless by next semester.