The Evolution of Ethernet

By Paula Bernier
Executive Editor, TMC

Networking technology has a way of disrupting business and displacing what came before it. But sometimes standards and solutions evolve along with the needs of the market. And that’s what seems to have happened with Ethernet.

A few years back some sources were predicting the demise of Ethernet technology. Instead, the use of Ethernet only continues to grow.

That’s because it has evolved to support ever larger communications connectivity. The arrival of Power over Ethernet means customers don’t require a separate power connection. And the new technology called FlexEthernet allows for flexible Ethernet connectivity between routers and optical transport equipment that is independent of the physical interfaces between the two devices.

Of course, Ethernet got its start as a local area network technology that worked at just 2.5 megabits per second. In the decades since then, Ethernet has expanded again and again to support double digit and then triple digit megabit per second connectivity, and then even gigabit level connectivity. In fact, a standard for 400 gigabit Ethernet is set to be ratified this year.

As for Power over Ethernet, it has been around for a long while with proprietary, and often non-interoperable, implementations from various vendors to power devices. But in June 2003, the IEEE (News - Alert) ratified the 802.3af PoE standard, which started the move to standard-compliant products in the PoE realm. And in 2009 the IEEE ratified the 802.3at PoE Standard, commonly known as PoE. That has allowed organizations to deploy equipment in locations like ceilings, kiosks, and walls that are difficult or too costly to have separate AC power installed.

As for FlexEthernet, it was defined by the OIF (News - Alert) Flex Ethernet Implementation Agreement. And FlexEthernet, or FlexE as it’s sometimes called, provides flexibility at the service layer. It does that by disassociating the client from the actual interface or server by introducing a new shim between the client and the PHY layer using a TDM frame structure.

FlexE was initially conceived to meet the demands of content providers for higher capacity and flexibility in data centers. Today there are three primary use cases for FlexE. Those include bonding (for example, 400GE supports today using existing 100GE modules), sub-rating (which matches the service rate to the available resource), and channelization (which aggregates low rate clients onto a group interface).