Devil's Advocate: Ethernet power

Power over Ethernet, power of Ethernet

Adding power distribution to Ethernet networks promises lots of improvements, says Martin Brampton. But what sort of step forward should we expect? It has been coming for some time but we can now celebrate the formal ratification of the 802.3af IEEE standard. That is the standard for providing electrical power through ordinary Ethernet cabling. Early efforts have, unfortunately, been proprietary. The standard is a major breakthrough. Seductive though it is, wireless will not solve all our problems, as I recently found with Bluetooth. It may help with the clutter of wires but power distribution is at least as pressing a problem. We also know from attempts at large wireless demonstrations that wireless is not yet a proven solution for high-density usage. Delivering modest amounts of electricity along the network cable will solve some problems very neatly. One pressing need is for IP phones to be powered from the network. Phones have always operated, at least to a basic level, on the power provided along the phone wiring. Without power over Ethernet IP phones could not do that. Moreover, the mess of power distribution on the average desk is a problem that has only been getting worse. The standard thirteen-amp plug is over-engineered for the small power demands of electronic equipment but alternatives such as IEC connectors (as used in almost all kettles and computers) are difficult to implement comprehensively. Solutions such as USB mobile phone chargers have been ad hoc responses to the problem. Another problem that will be solved by widespread availability of Ethernet connections that deliver power is that of international compatibility. At present, battery powered electronic devices need regular recharging, with the difficulty that power plugs vary greatly across the world. With 802.3af being an international standard such devices can be plugged into a network and powered anywhere in the world. All the same, it is too soon to be euphoric. The amount of power is strictly limited, mainly by the requirement that it be delivered in a form that does not present a safety hazard. The voltage is low to avoid the danger of electrical shock but that severely limits the power that can be carried by the standard conductors in Ethernet cabling. And that is a fundamental constraint of physics. The progress in the capability of, for example, laptop computers has come from a combination of less power hungry electronics and more powerful batteries. While many battery-operated devices could be powered by 802.3af now, all future developments will need to be on the power demand side of the equation. There is a major question as to which products vendors will choose to design with constrained power supplies in mind. There is also the question of how an agreed standard for power over Ethernet will affect network design. Initially, people will obviously take advantage of the standard’s compatibility with many existing cabling systems. But these are already under pressure from issues such as hot desking and very high densities of office occupation. Reliability is also an issue of increasing concern. Data networks have traditionally struggled to achieve service levels that would be considered acceptable for traditional telephony. One of the factors is power delivery to the many active devices that are traditionally scattered around networks. Any failure of power can take out a sizeable chunk of network. While effective power backup is achievable at central sites, it is much more difficult when distributed across the network. That suggests that further standards are needed to deliver larger amounts of power over backbone portions of networks. Here backwards compatibility will be much harder, partly because optical fibre plays a prominent part, and is inherently incapable of carrying much power. Optical power can be used to drive optical devices but is insufficient to support significant electronics. So, welcome as the new standard is, there will have to be a good deal of practical experimentation before we understand its full potential. There may need to be other developments too if we are to achieve networks that meet the standards demanded by the view of IT as a basic utility. ** Martin Brampton is a director and founder of Black Sheep Research (www.black-sheep-research.co.uk ), an independent consultancy providing research, writing and speaking services on a wide range of business and technology subjects. Martin was previously a director at Bloor Research, and has worked with IT as a user and analyst for over 20 years. He can be contacted at silicon@black-sheep-research.co.uk. For past Devil's Advocate columns see the links below, or type 'Devil' into our search engine.