Telecom Research Aims at Climate Protection

More users, more smartphones and tablets, and explosive growth in data-hungry apps are all adding up to massive new demand for network capacity in coming years. Data traffic has already grown larger than voice for many mobile operators in industrialised countries, a trend that can only become more pronounced as cell phone networks continue to converge with the Internet. In developed countries, operators project a doubling in data traffic each year for the foreseeable future, while income from that traffic is expected to grow by only 10 per cent annually.

Meeting the new demand while rolling out the next generation of cell phone services will push operators to invest hundreds of billions of dollars worldwide in buying the latest equipment and installing it in ever-denser configurations. And while they’re looking to pump up revenue by persuading cost-conscious consumers to pay extra for all these digital wonders, the phone companies are also tightening up their spending on operational costs wherever they can. That’s where KTH PhD candidate Sibel Tombaz’s research on network power saving can make a difference.

“Energy is a significant expense for mobile network operators,” Tombaz says. “There’s lots of engineering work going on to improve the efficiency of the hardware, like power amplifiers and cooling systems. But what about the design of the network itself? We can save a lot of energy by building networks that match power demand to user traffic. That means deploying the right number of large and small base stations so capacity is available when it’s needed, and turning off idle capacity when loads are low.”

The fundamental issue she studies has been around since the beginning of the cell phone era: Operators have to deploy enough base stations in a given area to handle all the users who want to make calls, but each extra antenna means added expense not only for the base station itself, but also for the mast or rooftop site, power, cooling and maintenance. Tombaz and her group are working on a tool that will allow network designers to make precise calculations of the energy costs and related carbon dioxide emissions of various configurations. An important element of the research is to build in data on local costs and available alternatives so the tool is relevant anywhere in the world.

“In urban areas of Europe or North America, energy may account for just 10 per cent of operating costs, but in rural parts of countries like India or Tanzania, energy might consume half of the operating cost because base stations aren’t connected to the electric grid,” she says. “I’m trying to show operators which energy parameters are the most important to modify so they can provide for projected increases in data traffic over several years while minimising energy consumption.”

One approach to cutting power consumption is to let the base stations talk to one another so the network as a whole adapts to traffic fluctuations. “You might need all your base stations in operation at dinnertime, but not in the middle of the night,” Tombaz explains. “So one station can tell the others ‘I’m going to sleep now. You handle my traffic and wake me up later.’”

Solving real problems

After receiving an undergraduate degree in Electrical and Electronics Engineering from Istanbul University in her native Turkey, Tombaz went on earn a Master’s in 2009, completing her thesis as a Visiting Scholar at the University of South Florida. 

Now, with the second year of her four-year PhD programme at KTH nearly finished, she finds a lot to like about the close co-operation between the university and the many successful telecommunication companies based near the School of Information and Communication Technology at KTH’s satellite campus in Kista, Stockholm’s high-tech suburb. There, KTH is in the heart of a focussed partnership between private companies, government and academia that aims to keep Sweden in the forefront of selected economic “clusters” that include wireless communication, medical technology, cleantech and nanotechnology. Kista is home to the giant Swedish telecom equipment maker Ericsson, and hosts research centres run by global leaders such as Microsoft, IBM and Nokia. 

Tombaz’s current research project, entitled “Energy Efficiency Measurement at the Network Level,” is sponsored by Ericsson. She says the close relationship with industry means her degree work will turn into practical instructions for operators to follow in measuring and modelling energy efficiency, making comparisons possible even between networks built to completely different technical specifications. 

“It’s great for us to have close relationships with companies like Ericsson, because we learn about the real problems facing the industry,” she says. “All respect for pure science, but the approach in this department is to see the challenges from the perspective of companies that have to survive in the real world. We don’t make up problems to study—we work on real problems.”

This academic-industrial symbiosis pays dividends both for degree candidates and for the companies that sponsor research. “These companies have their own R&D teams, but they also need to work with academics who can put a long-term effort into one topic,” Tombaz says. “When their own researchers go to work on a problem, they have to find a solution fast and move on to the next issue. But a PhD student might spend four years digging deep into a single issue. That looks good on your résumé when you get your degree—and when you start applying for jobs.”

But just landing a well-paying job isn’t all that matters to this ambitious 27-year-old—she’s driven by real concerns about a warming planet. “I do personal things like recycling and turning off lights, but just imagine if I can find a way to improve network energy efficiency by just a few percentage points,” she says. "Environmental awareness is a genuine part of Swedish culture. I’m not naïve, but I’d like to believe that companies like Ericsson reflect that.

“The ICT sector uses as much power as the entire airline industry, and this research means way more than anything I can do as an individual. I want to help change the world.”

By Kevin Billinghurst | kb2@kth.se