For all the variety in campus telecommunications-telephone, computer, security, fire alarm, cable television, building controls and more-the system uses just three types of signals: voice, data or video. Devising an effective way to distribute those signals is a major challenge for colleges and universities.
Two basic telecommunications problems plague virtually every campus: an existing infrastructure that needs upgrading, and a piecemeal collection of services in separate buildings that needs to be networked. These problems have been evolving over decades. First came telephone cable, then computer cable in the 1960s and '70s. When it came time to install new systems, whether it was communications technology, security or fire safety, contractors simply installed another set of wires.
The result: layers upon layers of mostly undocumented cable-some of it decades old-and decentralized systems that cannot always communicate with one another or be easily upgraded. Clearly, it is time for a change.
A flexible master plan Campus officials should create a master telecommunications plan that is flexible enough to accommodate future developments in technology. The plan should be revisited every five years or so.
For instance, a large university in the Northeast installed cable in 1986 for a voice-only infrastructure supporting telephone service. All the cable emanated from a single, central PBX switching station, and some of the cable runs were as long as 2,000 feet.
That worked fine for analog telephone connections. But as the university later discovered, this configuration prevented it from using a new technology: high-speed Integrated Service Digital Network (ISDN) lines, which allow simultaneous voice and modem use. Why? Because the maximum run for an ISDN line is 1,000 feet.
Infrastructure vs. platform Before planning a system, you should understand some of the basic issues. First, the difference between platform or protocol, and infrastructure: Infrastructure refers to the physical infrastructure that transmits voice, data or video signals; that is, the cable-whether it be copper, fiber optic or a combination. In a wireless system, infrastructure refers to the transceivers.
The platform or protocol refers to the electronics on each end of the cable, such as the Ethernet switches in a campus data network. In short, the platform refers to the active electronics that transmit signals through the cables, which are the passive, physical medium or infrastructure.
There are essentially four platform/protocol options: Ethernet, ATM, Sonet and Intranet. Which one you choose will affect the ability of your telecommunications system to support voice, data and video.
-Ethernet. A very popular choice for network connections on college and university campuses, the Ethernet uses Internet protocol (IP). It is an ideal choice for transmitting data. Nearly all commercial and academic LANs in the United States are Ethernet LANs. It has become popular because it is cost-effective, runs over Category 5 cabling, and supports large numbers of users. Its wide acceptance ensures that vendors will continue to work to enhance its functionality. However, it is a poor choice for transmitting voice and video.
-Asynchronous Transfer Mode (ATM). ATM was designed specifically to handle the problems of transmitting voice, data and video, and is a much better choice than Ethernet for future needs. If you were starting from scratch, ATM would be a good choice. But Ethernet is so entrenched that it does not make a lot of sense to take it out.
One solution is to deploy the Ethernet/IP protocol within each building, and tie all the buildings together with an ATM backbone. That still does not get video to the desktop, but it does position a college to extend the last portion of the ATM network if there is a future demand for it. It also supports the college's data networking needs. Schools can tie together separate campuses using ATM circuits.
-Sonet. Sonet is a high-speed fiber-optic-based networking protocol. Essentially a wide-area network (WAN), Sonet is a ring with two separate paths, which connect any number of individual nodes-such as separate campuses in a college or university system. Though expensive, it is fast and works well.
-Intranet. An alternative to Sonet, an Intranet connects separate campuses through a local Internet service provider. This alternative is usually cheaper than Sonet, but less reliable.
A college might consider using Ethernet in individual buildings tied together with an ATM backbone, and either ATM or Sonet to link separate campuses.
Infrastructure issues The building blocks of the telecommunications infrastructure are the cable distribution plants, trenches, conduit and cable that transmit the signals. Here are some of the issues to consider:
-Fiber-optic cable. Two types of fiber-optic cable are most commonly used in a telecommunications infrastructure: multi-mode and single-mode cable.
Multi-mode cable is made up of 62.5-micron glass strands and is used to send voice, data or video signals. It is mostly used for the Ethernet backbone and for video transmission. Single-mode fiber-optic cable is similar, but can send signals farther and faster, such as Gigaspeed Ethernet, ATM and Sonet applications.
Recently, some users have begun to phase out multi-mode cable because networks are becoming too fast for it. Nevertheless, a campus will want a combination of single- and multi-mode fiber.
-Copper. Copper cable is used for wiring within buildings, from the intermediate distribution frame (IDF) to individual workstations. Today, the most desirable standard of copper cable is enhanced Category 5, but a Category 6 standard is on its way. Category 5 cable is typically used for voice connections from the IDF, although it is more expensive than Category 3 cable, which has been used in the past.
-Transceivers for wireless systems. Will all our discussions about fiber-optic and copper cable become moot in 10 years when everything goes wireless? Probably not. It is unlikely that wireless networks will ever be fast, inexpensive or reliable enough to supplant wired systems. (Today the fastest wireless local-area network operates at 10 megabytes per second, while wired desktop systems operate 10 times faster.) But strategically placed transceivers can be used successfully to supplement cable connections.
For example, campuses can use transceivers to transmit signals from video cameras at a remote parking garage that is difficult to connect by cable. A school that cannot run cable under a public highway to reach part of its campus can use transceivers to link that area. Similarly, a school can supplement a PBX telephone with transceivers mounted in strategic locations around campus.
-Cabling topology. The design of a network's cabling topology, that is, the arrangement of distribution frames and cable, depends on the geography of the campus. A single, main distribution frame at the center of a star is the simplest cabling topology, but it is not always the most functional.
In the example of the university described above, a cabling topology of four interconnected wiring hubs, each with telephone service and situated so that no run was more than 1,000 feet, would have allowed the school to put its PBX switches anywhere and change over later to ISDN. Strategic planning is critical when designing an effective cabling topology for the long haul. The cabling may cost millions of dollars and provide service for 30 to 40 years.
-Duct banks and conduit. Few college campuses want to string their telecommunications cable from telephone poles. This means digging trenches and burying cable. That can be costly.
There are several alternatives to steel pipe for carrying cable. One is innerduct, corrugated plastic tube about two inches in diameter. It is cheap and fairly strong-most long distance carriers use it for their own networks. Another alternative, stronger than innerduct, is schedule 80 PVC pipe. It is very dense and provides more protection than innerduct.
Innerduct and PVC are buried in soil without concrete fill. Direct-bury cable, which uses no conduit, is the cheapest alternative and is nearly as strong as PVC pipe or innerduct.
But all of them can be torn out with a backhoe. For more security, you may want to bury steel pipe and fill the trench with cement. Which method you choose depends on your budget, the logistics of the job and the value of the data the cables are carrying.
You can minimize the risk of an outage by choosing a cabling topology that uses alternate routes. This can eliminate the need for expensive conduit installations. In any case, always lay additional empty conduit for future growth.
-Grounding and bonding. Fiber-optic cable does not conduct electricity, and grounding issues do not apply. Copper cable does, so there is a possibility of faults when the cabling system comes in contact with high voltage or lightning.
The National Electric Code (NEC) now requires that all cables entering a building to be grounded within 20 feet of the building entrance. You must consider this as you plan your infrastructure. In a retrofit, any cable vault that is going to be reused must be within 20 feet of the building entrance, or you must move it.
-Leak and fault detection. Do you need early warning systems to detect and pinpoint locations of water that is flooding a manhole or an electrical interruption? That depends on the type of traffic and the value of the data traveling on the network. A long-distance telephone service provider with a 500-mile long cable will want to pinpoint the location of a break. But for most colleges and universities, there is little practical value in installing expensive leak- and fault-detection systems on cable duct banks. You will remedy the failure by replacing the entire length regardless of the location of the fault.
-Inside the building. Thus far, the discussion has focused on the system of main cable distribution frames, conduit and cables that interconnect points on the campus. Once inside a building, intermediate distribution frames (IDF), risers, telephone closets and above-ceiling or under-floor-conduit take over. This is where teamwork with the architect early in the planning phase is essential.
You should have riser shafts with plenty of space for future cabling on every floor, or on alternating floors. Spaces and pathways must be available to pull cable above suspended ceilings or through the slab, from the IDF to the workstation, classroom or residence-hall room.
A retrofit is more complicated. What if you have a classroom building with sheetrock ceilings and concrete slab floors? What if it is a landmark? These and other issues may come into play.
Start over? For many colleges and universities with old telecommunications infrastructures and an ample budget, the logical solution is to remove the old infrastructure and start over. That is exactly what some are doing. These schools are installing a modern infrastructure parallel to the old one, cutting over to the new systems in well-orchestrated phases and removing the old systems.
Equally important, they are including provisions for growth and change from the ground up. This is very expensive, to be sure, but so is the cost of being unable to deploy new technology, or deploying it and finding it is unreliable.
In addition to providing basic services through a campus-wide voice, data and video network, the well-wired college or university can take creative advantage of new technology in many areas.
-Instruction. With a laptop, a professor at a lectern can control a ceiling-mounted projector to present any available electronic material, including information from the Internet. Students can retrieve lecture material on network archives. Video cameras in the classroom allow students to "attend" class through a residence-hall-room computer. Video teleconferencing can connect other remote users.
-Library. Students anywhere on campus can access unlimited "copies" of any electronically stored text, audiotape or videotape. Librarians can search for, shelve and inventory bar-coded materials using a portable computer and wireless connection to the electronic card catalog.
-Administration. Students can apply for admission online at the university's web site. Students register for classes online rather than on a long line, and can view course descriptions and faculty evaluations on their computers. They check their grades online, as well as the status of accounts with the university. The school stores student information on "smart" ID cards that can control and monitor building access, food service, long distance calling, spending for public transportation, and use of debit cards at local businesses. As an additional barrier to cheating, test proctors can authenticate the identity of students.
-Web access. A well-wired university will have a data network port for every bed in every residence hall and multiple ports in public areas, including residence hall lounges, classrooms, labs, the student union and the library. A university can buy bulk bandwidth with a T-3 connection (45 mbs) to an Internet service provider and offer students and faculty high-speed access to the World Wide Web.
