Ethernet celebrates 40 years

Happy birthday, Ethernet: 40 years ago, networking got a lot easier

May 22 marks the 40th anniversary of Ethernet, the nearly ubiquitous local area networking technology, at least if you measure its birth from Robert Metcalfe’s seminal 1973 memo describing the scheme.

Not everyone does. David Boggs, Metcalfe’s collaborator on the networking project at the Xerox Palo Alto Research Center, dates Ethernet’s birth from its first functioning system, which was six months later.

Either way, Ethernet has had a long run, emerging as the dominant LAN standard while expanding its capacity, range and functionality.

“The Ethernet invented 40 years ago is quite different from the Ethernet that is running today,” said John Hawkins, the Carrier Ethernet expert at the optical networking company Ciena. And yet, there is a lot that remains unchanged. The frame format is largely the same today, Hawkins said, its ease of use and plug-and-play capability is the same, as is its ability to fit easily into multiple network niches.

The ability for any connected device on the Ethernet wire or cable to communicate with any other connected device without modifications to the devices or the network gave it an advantage over competing technologies such as Token Ring and Fiber Distributed Data Interface (FDDI). “Ethernet won the early wars based on its simplicity and cost points,” Hawkins said. “It enjoyed a great cost advantage.”

Adopted as the IEEE 802.3 standard in 1985, it has enjoyed an economy of scale that has kept it affordable. It also has benefited from the fact that while it has advanced in speed and distance — from a top speed of 10 megabits/sec and a maximum wiring run measured in tens of meters to gigabit speeds and spans of tens of kilometers — it has remained largely backwards compatible. Older 10 megabit/sec devices will work fine on Gigabit Ethernet, just slower.

As originally described, Ethernet was a simple physical and data layer scheme to allow devices to communicate over a common wire (initially coaxial cable). Data is transmitted in variable-length frames containing a source and destination address, and each device on the network checks the destination address of each frame. If the frame is not addressed to the device, it is ignored. Each device listens to make sure its network segment is clear of traffic before transmitting, and when frames collide in a segment, the devices stop transmitting, wait a bit, and then check to see if the segment is clear again before resuming.

Ethernet networking has been extended beyond its initial limits through the use of repeaters in the physical layer (Layer 1), bridges in the data layer (Layer 2), and switches. Switches are Layer 2 devices that forward frames only to destination ports, effectively creating a dedicated segment for each networked device, reducing the competition of traffic on the segment.

Originally working over coax in the 1970s and 1980s, Ethernet today typically runs over CAT5 and CAT6 copper and fiber optic cable.

As the term “Carrier Ethernet” implies, Ethernet is expanding beyond the enterprise and into metro- and wide-area networking, and Hawkins said he expects to see the trend continue.

“There will be more capacity, more bandwidth and applications flowing over Ethernet,” he said. “Further expansion will always be on the horizon.”

About the Author

William Jackson is freelance writer and the author of the CyberEye blog.

Reader Comments

Please post your comments here. Comments are moderated, so they may not appear immediately after submitting. We will not post comments that we consider abusive or off-topic.

Please type the letters/numbers you see above