1 August 2006
Ethernet arose from Aloha Net
By John Rinaldi and Perry Marshall
Ethernet is the worldwide de facto standard for linking computers together.
Ethernet connects hundreds of millions of computers and smart devices across buildings, campuses, cities, and countries. Cables and hardware are widely available and inexpensive (dirt cheap in the case of ordinary office-grade products), and software is written for almost every computing platform.
Ethernet is now a hot topic in automation, where industry-specific networks have dominated: Profibus, DeviceNet, Modbus, Modbus Plus, Remote I/O, Genius I/O, Data Highway Plus, Foundation fieldbus, and numerous serial protocols over the electrical standards of EIA RS-232, RS-422, and RS-485.
In some cases, Ethernet is displacing these networks. In nearly all cases, Ethernet works in demanding installations alongside them.
Ethernet originated at Xerox Palo Alto Research Center in the mid-1970s. The basic philosophy was any station could send a message at any time, and the recipient had to acknowledge successful receipt of the message.
It was successful, and in 1980, the DIX Consortium (Digital Equipment Corp., Intel, and Xerox) formed and issued a specification. The Institute of Electrical and Electronics Engineers (IEEE) took it over and in 1983 issued the Carrier Sense, Multiple Access/Collision Detect (CSMA/CD) specification, their stamp of approval on the technology.
Ethernet has since evolved under IEEE to encompass a variety of standards for copper, fiber, and wireless transmission at multiple data rates.
Ethernet is an excellent transmission medium for data, but by itself falls short of offering a complete solution. A network protocol is also necessary to make it truly useful, and what has evolved alongside of Ethernet is TCP/IP.
From both a historical view as well as in today’s industrial world, the TCP/IP plus Ethernet marriage is a key combination, without which, neither would have survived.
The many formats and nomenclature of Ethernet cabling is a rather unfriendly shorthand terminology. IEEE’s Ethernet naming convention works like this:
The first number (10, 100, 1000) indicates the transmission speed in megabits per second.
The second term indicates transmission type: BASE = baseband, BROAD = broadband.
The last number indicates segment length. A 5 means a 500-meter segment length from original Thicknet.
In the newer standards, IEEE used letters rather than numbers. The T in 10BASE-T means Unshielded Twisted- Pair cables. The T4 in 100BASE-T4 indicates four pairs of Unshielded Twisted-Pair cables.
Packet communications key
During the 1970s, a Harvard graduate student named Robert Metcalfe read a paper about something called Aloha Net. It was a radio system used in the Hawaiian Islands to send small messages, also called data packets, between islands.
A key feature of this network was anyone could send messages at any time. If no acknowledgement as to receipt came, the message would retransmit.
Metcalfe reasoned correctly that with some mathematical enhancements to the system, the efficiency of the Aloha Net could drastically improve.
The packet communications network he designed became the worldwide standard we know today as Ethernet. It is also IEEE Standard 802.3; it retains today the elegance and simplicity of the original Aloha Net.
Above is a simplified diagram of the contents of an Ethernet packet. It contains two address fields, some data, and a field that verifies correct reception of the packet.
The first two fields are unique 48-bit addresses of the sending computer and the destination computer. These addresses are not the familiar 192.168.0.10 type addresses we often see but are addresses assigned by the manufacturers of the physical Ethernet cards.
Every manufacturer producing Ethernet hardware has, by assignment, a series of 48-bit addresses.
Known as the Media Access Control (MAC) address, the manufacturers of Ethernet interface cards must ensure they use only the addresses in their series and use it only once. That way, no two computers in the world can have the same address.
ABOUT THE AUTHORS
John Rinaldi (jsr@rtaautomation. com) is the president of Real Time Automation. Perry Marshall (info@ perrymarshall.com) is a networking consultant. They wrote the handbook Industrial Ethernet, ISA Press, 2005.
Baseband is a transmission method in which the entire bandwidth (the rate at which information travels through a network connection) is used to transmit just one signal. Baseband is a cheaper method than broadband and is typically used for shorter distances of transmission.
Broadband refers to data transmission where multiple pieces of data transmit simultaneously to increase the effective rate of transmission. In network engineering, it is where two or more signals share a medium.