An engineer named Bob Metcalfe conceived the idea of Ethernet in 1973, while working for the Xerox research center in Palo Alto, California. His fundamental invention was the CSMA/CD method of channel arbitration, allowing multiple devices to share a common channel of communication while recovering gracefully from inevitable “collisions.” In Metcalfe’s vision, all of the “network intelligence” would be built directly into “controller” devices situated between the DTE devices (computers, terminals, printers, etc.) and a completely passive coaxial cable network. Unlike some other networks in operation at the time, Metcalfe’s did not rely on additional devices to help coordinate communications between DTE devices. The coaxial cable linking DTE devices together would be completely passive and “dumb,” performing no task but the conduction of broadcast signals between all devices. In that sense, it served the same purpose as the “luminiferous ether” once believed to fill empty space: conducting electromagnetic waves between separated points.
The CSMA/CD (“Carrier Sense Multiple Access with Collision Detection”) method of bus arbitration works by giving each Ethernet device the ability to sense an idle channel as well as sense if it happens to “collide” with another transmitting device. In the event of a collision, the colliding devices both cease transmission, and set random time-delays to wait before re-transmission. The individual time delays are randomized to decrease the probability that a re-collision between the same devices will occur after the wait. This strategy is analogous to several peers in one group holding a conversation, where all people involved are equally free to begin speaking, and equally deferential to their peers if ever two or more accidently begin speaking at the same time. Occasional collisions are perfectly normal in an Ethernet network, and should not be taken as an indication of trouble unless their frequency becomes severe.
Metcalfe’s original network design operated at a data rate of 2.94 Mbps, impressive for its time. By 1980, the three American computer companies DEC (Digital Equipment Corporation), Intel, and Xerox had collaborated to revise the Ethernet design to a speed of 10 Mbps, and released a standard called the DIX Ethernet standard (the acronym “DIX” representing the first letter of each company’s name). Later, the IEEE Local and Metropolitan Networks Standards Committee codified the DIX Ethernet standard under the numeric label 802.3. At the present time there exist many “supplemental” standards underneath the basic 802.3 definition, a few of them listed here:
• 802.3a-1985 10BASE2 “thin” Ethernet
• 802.3d-1987 FOIRL fiber-optic link
• 802.3i-1990 10BASE-T twisted-pair cable Ethernet
• 802.3u-1995 100BASE-T “Fast” Ethernet and Auto-Negotiation
• 802.3x-1997 Full-Duplex standard
• 802.3ab-1999 1000BASE-T “Gigabit” Ethernet over twisted-pair cable
The IEEE 802.3 standard is limited to layers 1 and 2 of the OSI Reference Model: the “Physical” and “Data link” layers. In the physical layer (1), the various supplements describe all the different ways in which bits are electrically or optically represented, as well as permissible cable and connector types. In the data link layer (2), the IEEE standard describes how devices are addressed (each one with a unique identifier known as a MAC address, consisting of a 48-bit binary number usually divided into six bytes, each byte written as a two-character hexadecimal number), and also how data frames are organized for Ethernet transmissions.