Home

Teaching

Service

Curriculum Vita

Courses on the Web

Lane Report Columns

College Technology

Search Resources

Technology Links

Music Links

The Music In Me

School of Journalism
and Telecommunications

College of Communications
and Information Studies

Ethernet Bridges the Gap
The newly-developed network offers numerous advantages

Those of you who work at large institutions like the University of Kentucky or corporations with multiple locations on a campus know that a considerable amount of time, effort, and money is devoted to operating telecommunications networks over a large area. Generally, the network layout consists of a number of local area networks (LANs) within buildings that are then connected to each other by another network called a "backbone." Depending on the size of the institution, there may be intervening layers of sub-networks between the LANs and the backbone. Beginning with the LANs and going up to the backbone, each higher level of the network carries greater volumes of data as more and more data streams are combined. This is what is known as a hierarchical network.

UK's network is a typical example of a fairly advanced campus network. Most of the buildings on the campus now have an installed base of Ethernet LANs that operate at 10 Mbps. Some locations are fortunate enough to be running at 100 Mbps, and virtually all new LAN installations are capable of the 100 Mbps Ethernet standard. The campus backbone is an Asynchronous Transfer Mode (ATM) network. The sub-nets of the network operate at the OC-3 standard, 155 Mbps, and the main backbone runs at the OC-12 standard, which is 622 Mbps. OC stands for optical carrier, so naturally the backbone physically consists of fiber optic cables, while the LANs are constructed of Category 5 twisted pair wires. UK's network serves the needs of the university admirably, but like all current technology it is becoming obsolete as you read this.

ATM came to prominence in the early 90s and has been the standard for comparison for most of the decade. In its original conception, it was envisioned as a standard that would blur the lines between local area networks and wide area networks. This idea, however, rested on the notion that ATM would go right to the desktop computer via something called ATM LANE (local area network emulation). That development, for various reasons, has not really come to pass. Nowadays, the challenger to ATM as a backbone protocol is something called Gigabit Ethernet, which has been in development for the last two or three years. Of course, the most apparent benefit is the gigabit part -- this type of backbone network runs at 1000 Mbps. (By the way, the "M" in "Mbps" stands for "mega," which means million. That means that a gigabit transmission is one billion bits per second!)

Its most distinct advantage over ATM, however, springs from the fact that it uses the same Ethernet 802.3 standard that is the basis of 10/100 Mbps local area networks, which exist in large numbers at UK and elsewhere. Perhaps a little explanation is in order. One type of device for connecting one network to another is called a bridge. A bridge is a comparatively simple and inexpensive device because it connects two or more networks that use the same protocol. A router is another kind of connection device. Routers connect dissimilar types of networks. They have to perform translation and conversion for different protocols, packet sizes, and transmission speeds. Necessarily, a router is more complex and more expensive than a bridge. Therefore, because the LANs and the backbone use the same protocol, the Gigabit Ethernet network is easier to administer and cheaper to build and operate.

If your corporation or institution is considering scrapping that old token ring network or that FDDI (fiber distributed data interface) network anytime soon, and you already have an installed base of Ethernet LANs, then there's really only one way to go -- gigabit.