Energy Consumption in Wireless Networks: The Big Picture

By | November 6, 2013

Green Energy

I recently came across a presentation on advanced antenna systems with the statement: “advanced antenna systems for power consumption savings not for capacity.” I was very intrigued for a couple of reasons. The first is how much of a problem is power consumption in wireless networks is. The second is that I recalled a conversation I had over 14 years ago with a colleague at Metawave prior to joining them. He said that they were approaching smart antenna systems from the perspective of capacity and not coverage. Back then, the nascent technology was traditionally targeted at improving coverage which was the reason why these systems failed to get traction in the market. So, today, we are changing the pitch for these systems from a capacity focus to a power savings focus. But will that make them more attractive? How much of a problem power consumption is?

Let’s look at some back of the envelop numbers to frame the issue. A base station site consumes between 1000 – 2000 W (and often more), depending on a number of factors such as the number of radios, frequency channels, and traffic load.  For a typical US operator with about 50,000 sites, that over $60mil a year in operational expenses just to power the radio access network (RAN). The RAN accounts for about 70% – 80% of the total power, the rest is consumed by the core network. The total is then over $90mil – and I think this is a conservative number.

Cost of powering the RAN
BTS Power consumption

2,000

W
Number of sites

50,000

 
Energy consumed

    100

MWh
Price of electricity

0.07

$/kWh
Total

  61,320,000

$

Taking a macro view for a top-bottom approach, the telecom industry accounts for over 1% of the total world energy consumption. I found the table below shows the energy consumption of some leading telecom companies in the world from 2008. Today, Verizon’s total energy consumption is on the order of 10.5 TWh up from 8.9 TWh in 2008 – of course, this is an entire company’s power consumption, wireless and wireline businesses included. Verizon’s annual operating budget is on the order of $46bil. So power consumption in the RAN accounts for about a fraction of 1% of the total operating budget. The question is then: is power consumption a significant issue to sway operator’s technology roadmap?

Source: Emerson Network Power

Source: Emerson Network Power

Verizon Electricity Consumption (Source: Verizon)
Year 2009 2010 2011 2012 % change
Electricity (TWh) 10.27 10.24 10 10.47 1.90%

There are a few favorite topics in the wireless industry that everyone likes to talk about such as capacity and stale ARPUs. But green energy in wireless networks is a much less ‘sexy’ topic that is only discussed in few focused forums without much media attention and it has not been one of the top priorities for CTOs despite limited projects to use renewable energy to power base stations.

As we move to LTE, we can expect an increase in energy consumption because LTE radios are less efficient than 3G radios due to the OFDM physical layer and requires more radios for MIMO. Radios account for anywhere between 40-80% of the base station total power consumption.  For this reason there has been a fair bit of work on improving the efficiency of power amplifiers. There are other techniques also used to reduce overall power consumption like the adoption of remote radios. So while demand on energy increases, there are new techniques being introduced to keep energy consumption in check.

Going back to advanced antenna systems, this is a further evolution where the remote radio is distributed across the antenna elements to create beams that can be changed in orientation and focus to meet base station performance requirements and optimize for energy consumption. But will the operators adopt such solution? Should investors invest in companies targeting green products for wireless networks? I think framing the question simply based on power consumption will not be enough to sway operators, but there has to be a real value in cost-performance trade off compelling enough for their adoption.