•  
• EDN Europe MAGAZINE
•  
• Free SUBSCRIPTION
• Subscription Renewal
• View Media Kit

• EDN
• EDN Australia
• EDN Asia
• EDN China
• EDN Japan

EDN.COMMENT: “Greening” the Internet

Agroup of 15 organisations drawn from across the IT and networking industries, and inspired by a study initially conducted at Bell Labs, comprise the founding members of the Green Touch alliance. Over the next five years, Green Touch aspires to developing the enabling technology to build a network infrastructure that is 1000-times more energy-efficient than today’s technology permits.

Alcatel-Lucent CEO Ben Verwaayn asked a Bell Labs research team, in mid- 2009, to consider the margins by which energy efficiency in telecoms and datacomms networks might be improved. Alcatel then drew in a core group of industry manufacturers, suppliers, academics and research organisations to form Green Touch.

Its first steps will be to formalise the consortium’s structure and set out a “roadmap” of technology targets, over the early months of 2010. After five years, Verwaayn says, the deliverables should be enabling technology for affordable, backward-compatible network products that are so much more energy-efficient, that their adoption will be attractive to service providers for that reason alone.

The concept that the energy levels we use today to represent information —in this context, information in transit—are many orders of magnitude greater than the fundamental physical levels that could in principle do the job, is hardly revelatory. And, indeed, the presentations launching Green Touch included passing references to the work of Claude Shannon. Shannon’s exposition is usually structured to define the maximum capacity or bandwidth of a channel, but it also tells us about bandwidth— or energy—per bit. Where, however, does the figure of 1000x come from, and how is it to be achieved?

Verwaayn’s scientists and engineers appear to have done rather more than simply making educated guesses. They seem to have attempted to make a weighted average assessment—of “wasted” energy—from a broad range of technologies employed in the network. Some of the numbers are very large; you can, for example, say that in a wireless link, all of the radiated energy that is not intercepted by the receiver is “wasted”. But not all of the inefficiency is available to reclaim; you might apply the most sophisticated coding to maximise the capacity of a channel—but the energy needed in the codecs to implement that coding will eat into any gains you make. Thinking through the entire network in this way, factoring in all the trade-offs, the Bell Labs team arrived at a 10,000x factor. Scaling that back to 1000x yields a target that might be both useful and achievable.

Less evident, perhaps unsurprisingly at this stage, is much guidance on the “how” of the project. What will green network technologies actually comprise? While basic principles confirm that there are big energy savings to be made, how many of them would not be made under pressure of market forces, as time goes by, in any case? The sceptical observer might view the whole exercise as a means of gaining some positive headlines in a slack business period; you cannot, today, ever be “too green”. That would probably be overly cynical but, if nothing else, this project will face an acute version of the classic dilemma faced by such grand plans. Which is to maintain momentum, minimise bureaucratic overhead, and keep sight of the original objectives—while actually achieving meaningful technological progress. Not to mention allowing participants to take away something of value in return for their efforts, at the same time as keeping the collaboration pre-competitive.