Louis (Sam) Samuel, executive director, Bell Labs UK

By Sam Samuel

Sam Samuel, executive director of Bell Labs UK talks about reducing energy consumption in the future.

Energy is one of the big problems of the future for the industry. We’ve seen a number of studies on internet traffic growth, so one of the first things we did was to confirm that growth rate, because what happens next depends on the assuredness of that data. Then the next thing we looked at was what is composing that growth in terms of energy consumption.

It transpires that the core network piece is not that bad. But when you take into account wireless data things become really problematic. Especially if wireless data is provided in an unchecked way.

Wireless will overtake the rest of the internet in terms of energy consumption sometime between 2011 and 2013. This takes into account new technologies such as LTE, and means that even at best, using technologies we know about today, we won’t be able to keep pace with energy consumption.

So now we have to go back and challenge established wisdom. We need to work out what we can do with physics to reduce the amount of energy we need to transmit information across the wireless channel. We’ve got MIMO, beam-forming and a combination of the two, among other solutions, but none of them offer the kind of energy reductions you can get simply by reducing the size of the cell.

But that’s counter intuitive, of course: If you go from having one large cell to a number of smaller ones, then you are using more processing, which in turn is using more energy. But from here we have to go back and reassess exactly how we do the circuits. Can we go back and change them to give the power to performance ratio we’re looking for? You have to take everything you know to be true, throw it away and start again.

People are going to have doubts about this from a practical point of view. But if you take a city like London, for example, cell sizes are already shrinking. So you’re targeting areas where this is happening anyway. But the method of execution has to be far better than it is today; you have to change your way of thinking. Femtocells are in the right ballpark but you have to reconsider how they are powered, how you get backhaul to them, and how the circuits are made up.

The issue then becomes that, once you’ve sorted the issue for wireless, you then have to sort it for fixed access, because the femtocell is attached to the fixed access network. Then when you’ve figured that one out the next one back it is the core network—the bit we acknowledge as being pretty efficient already—but it still needs to be made more efficient.

The most interesting thing here is that the circuits we employ in the telecoms industry in our equipment are the same circuits used elsewhere, so this has repercussions universally. If we address our networks, it starts to address other business cases that don’t yet exist because you start to change the operating point of the network. When a network is more energy efficient your cost point shifts dramatically, but it can affect adjacent areas, too.

Add your own predictions in the comments box below or read those of other industry leaders in
2020 Vision