Nokia Bell Labs and KDDI Research team up on energy efficient cellular base station

Base station on the mountain in the sunny day

With a new method that demonstrates how to significantly reduce power consumption in 5G base stations without compromising network speed or throughput, Nokia Bell Labs and KDDI Research are auguring a more sustainable future for mobile networks.

Anytime you stream a video, join a call or use an app on the move, you’re relying on a nearby base station, which is the radio equipment that connects phones and connected devices to the mobile network.

The age of the AI supercycle will inevitably create a continued spike in mobile data demands. The increasing use of video, cloud services and AI-native systems require more network capacity with inherent burstiness. That will put a strain on base stations and contribute significantly to higher energy costs and carbon emissions. 

That’s why Nokia Bell Labs and KDDI Research are putting energy efficiency at the heart of their 6G work. The next generation of mobile technology needs to be faster and more capable, while also using significantly less power per bit of data delivered.

What our partnership is building

KDDI Research and Nokia Bell Labs have once again combined their strengths — KDDI’s operator insight and real deployment and Bell Labs’ deep wireless research — to explore how future networks can be both high performance and energy efficient. 

The latest milestone is a technique we call Intelligent 4D Resource Optimization. Put simply, it helps a base station “right-size” how it uses radio resources moment by moment depending on the amount of traffic it needs to send, so it doesn’t burn energy when it doesn’t need to. The “4D” refers to four basic knobs a base station can adjust: how long it transmits (time), how much spectrum it uses (frequency), how many antennas it activates (space) and how strongly it transmits (power).

In greater detail, the “time” element refers to sending data in a smarter rhythm so the equipment can rest for longer periods between bursts. The “frequency” is about using the right amount of spectrum needed for current traffic. “Space” refers to turning on only as many antenna elements as the situation requires and powering down the rest of the time. And “power” is about transmitting at a lower power when there is lesser traffic.

The key is coordination. Instead of optimizing each “knob” separately, the system balances them together in real time. That means it can sometimes use a little more of one resource to save more energy elsewhere, all while keeping the user experience strong.

What we demonstrated

We ran the proof of concept of this approach and then validated it in a trial during the month of March at Nokia Bell Labs headquarters in Murray Hill, New Jersey. 

The test environment was designed to reflect the kinds of conditions base stations face in real 5G networks. With AI-powered Intelligent 4D Resource Optimization, we observed up to 40% lower power consumption at the same throughput levels. We also demonstrated that higher throughput was possible without increasing power consumption compared with today’s 5G base-station equipment—to the tune of a 4x improvement in energy efficiency.

Several effective energy saving solutions have been deployed but this innovation takes energy efficiency to the next level.

What’s Next?

On April 23, Nokia Bell Labs and KDDI Research signed a new joint research agreement to push this work further and explore additional ways to reduce network energy use. Next, we’ll look beyond a single base station—studying how coordination across multiple sites and across different frequency bands can amplify savings at the network level. We also plan to share learnings with the broader industry as 6G standards take shape in 3GPP, helping ensure that energy efficiency is designed in from the start—not added as an afterthought.
 

Jinfeng Du

About Jinfeng Du

Jinfeng received his Bachelor degree from USTC, China, and his Master, Licentiate, and PhD degrees from KTH, Sweden, all in Electrical Engineering. Before joining Bell Labs in NJ, Jinfeng spent two years at MIT as a postdoctoral researcher. Jinfeng's research interests are in the general area of wireless communications, especially in communication theory, multi-user information theory, radio systems design and evaluation, millimeter wave propagation and channel modeling. Our department in Radio Systems Research Lab is to develop novel, superior, and disruptive technologies for wireless communication and sensing focusing on the physical and medium access layers, radio resource management, and system design. We are devoted to establishing fundamentals as well as to developing game-changing, high-impact technologies and concepts for radio connectivity in an ever-widening range of frequency bands.

Prasanth Ananth

About Prasanth Ananth

Prasanth heads the RAN Platforms Research department in Nokia Bell Labs. He leads a multi-disciplinary team which conducts applied research and prototyping for wireless sensing, AI-RAN, and energy efficiency. Previously, he setup and led the Aerial Robotics research team, developing autonomous indoor drones with edge intelligence, and founded the AIMS venture, where he served as the CTO. Prior to that, he led research on access networks and chaired global working groups in the OpenGen 5G UAS and GreenTouch consortiums. A Ph.D. graduate from the University of Maryland with a Postdoc at UT Austin, Prasanth holds 8 patents, 28 publications, and is recognized as a Nokia Distinguished Member of Technical Staff (DMTS).

Article tags