Understanding Ethernet

Electrical Apparatus, Mar 2002 by Tryling, David P

Looking at the history and the basics of this 'hot' communication medium

COMMUNICATION, DATA EXCHANGE, CONnectivity. These are some of the current topics that are changing the face of automation. Connecting PLC's (programmable logic control), displays, distributing I/O-all of these involve communication methods.

These communication methods have become the pathways to act as one in many automation applications. None of these currently is "hotter" than the growth of industrial Ethernet. It seems that everywhere we turn we see new products that are Ethernet-compatible. As maintenance and automation professionals, we look to Ethernet as one of the tools to make our operations more productive and our work easier.

History

Ethernet was developed in the 1970's by Dr. Robert Metcalfe at the Xerox Palo Alto Research Center. Its original purpose was to connect PC's or workstations together to form the "office of the future." This experimental system ran at the then-blazing speed of three megabits per second.

The first formal specifications for Ethernet were published in 1980 by a well-known group: Dec, Intel, and Xerox. Xerox had the patent, but the consortium created an open standard available to everyone. This became known as the DIX standard.

By this time, Ethernet had advanced to a speed of 10 megabits per second. Shortly after the DIX standard, the technology was accepted by the LAN standards committee for IEEE (Institute of Electrical and Electronics Engineers) and was published in 1985, becoming IEEE 802. This first IEEE 802 standard based on the original DIX standard.

IEEE 802 has been the guiding reference for Ethernet equipment since 1985. Following supportive practices, IEEE has updated 802 frequently to allow for technological advances. Currently, Ethernet is covered up to 100 megabits per second.

Ethernet defined

Like most complex technology, there is no short and concise definition of Ethernet other than a method of communication between PC's on a local area network, or LAN. After stating this, more details are necessary and the IEEE 802 becomes the guide. The overall standard defines a method of communication between independently operating computers where each has access to a common structure for communication.

The IEEE 802.3 standard was structured based on the Open Systems Interconnection (OSI) Reference Model. This structure model was developed in 1978 by the International Organization for Standardization (ISO). The ISO sets numerous vendor-neutral standards for technologyrelated items and developed the OSI model to assist in standardizing network topics.

Therefore IEEE 802.3 based on the OSI model breaks down Ethernet into seven layers. The first layer is the physical layer and is the most basic part of Ethernet. The whole concept behind Ethernet was that multiple computers share a passive connection. This passive connection is the cabling and other hardware devices as defined in the physical layer.

Passive means that there is no controlling device operating this medium and it is essentially nothing more than a pipe. Each computer listens on the pathway for communications directed to it. If the computer has information to send, it listens for a clear space and transmits. The Data Link Layer is the second layer in the OSI model and arbitrates the control of these transmissions. This provides for a method of arbitration of signals on the network and detection of collisions.

The process for this is called CSMA/CD, or Carrier Sense Multiple Access with Collision Detection, where "carrier sense" means listening on the network. If there is an opening on the network, any computer can "talk." This is called "multiple access." "Collision detection" is the process of determining if two computers are speaking at the same time. Putting this together, we have the CSMA/CD protocol, which becomes the heart of controlling communication across the Ethernet medium.

The OSI model defines the Data Link Layer in two pieces-the Medium Access Control, or MAC, and the Logical Link Control, or LLC. MAC controls access to the Physical Layer and the LLC is for data identification. IEEE 802.3 defines the CSMA/CD process that the Physical Layer and the Data Link Layer control. Layers 3 through 7 of the OSI model involve the organization of data and communication with application software in the PC and are not involved with negotiating communication on the network. These upper layers are left to the software vendors for application development.

The physical layer or, in our world, the cabling is defined by the descriptor used to identify the network. The standard describes these and provides for multiple methods of cabling. Some examples are 10-base-2, 10-base-F, and 10-- base-T, where "10" equals the frequency in megahertz of the network, base is short for baseband (which is a synonym for broadband), and the final modifier describes the type of cable. The number 2 is a coax cable, F means fiber-optic cabling, and T represents unshielded twisted pair cabling. Each type of cabling has its benefits as well as its downsides. Currently, the most common is 10-base-T or 100-- base-T.