Power over Ethernet (POE) is a seemingly innocuous innovation that will have a huge impact on the way we design, purchase, install, and manage our networking infrastructure. In a nutshell, POE allows us to power small network-ready devices without a separate power supply. Instead, an injector (which is either external or built into a switch) provides power over unused wires in the Ethernet twisted-pair cable.
POE has a lot going for it, including accessibility where there are no nearby AC outlets. It can also eliminate many of those unsightly cords and transformers as well as the need for having an uninterruptible power supply (UPS) at the injector source (one UPS on a POE-ready switch can cover dozens of devices). With the advent of the 802.3af standard, POE is taking off like a rocket.
There are a huge number of components available for POE, from chips to injectors to RJ-45 jacks with integrated Web servers. And weve only seen the tip of the iceberg for plug-and-play products. A sampling of POE-ready devices include VOIP phones, 802.11 access points and bridges, Web cams, and card access systems. There is even a clock that synchronizes to the National Institute of Standards and Technology (NIST) standard and provides local-time services via Telnet and a low-power, XP-based, flat-panel computer.
In our quest to lower total cost of ownership (TCO) and reduce the cost of our physical infrastructure, POE seems ideal. But there are some potential pitfalls to be aware of.
For one, a proliferation of Ethernet "micro-nodes" will require more switch ports to support them and perhaps upgrades to support POE. Historically, weve gone from shared cable (the original Ethernet), to hubs, to switches and hubs, to a switch port for every user. Now imagine a ten-to-one ratio of switch ports-to-users in the not-so-distant future.
