It’s breathtaking, seeing all human activities shine through a satellite image. But what we often forget is that behind the beauty lies massive energy consumption.
Today, the telecommunication industry accounts for many human activities, claiming 3% of global energy consumption. 5G and automation have enabled a more spectrum-efficient network, but the exponential growth in demand outweighs the energy reduction. Telco vendors tackle the challenge on a component level, but we’re viewing the problem with a broader lens, looking at mobile operator networks at the system level. With a holistic perspective, we aim to raise the industry benchmark in network energy management.
Here at Rakuten Symphony, our vision is to disrupt and democratize the industry by leveraging our full potential of cloud, data and Artificial Intelligence (AI). One part of that vision is disaggregating monolithic and energy-consuming elements into smaller components — the heart of Open Radio Access Network (Open RAN). To begin, Rakuten Symphony CEO and Rakuten Mobile CTO Tareq Amin spearheaded the world’s first fully virtual cloud-native mobile network, achieving a 96% coverage ratio in Japan. Now, we are leaping forward, accelerating the deployment of Open RAN mobile networks globally.
On this journey, we apply cloud technologies in our large data center architecture, regional data architecture and edge data center that allows an agile network. We relentlessly strive to provide an energy-efficient, intelligent and secure operation. And we are just beginning.
With necessary disaggregation of diverse RANs in 5G, you might ask, “isn’t that going to consume more energy?” The answer is no, as the application of intelligence leads to better performance on a grand scale. The current approach to energy management in traditional networks is equivalent to keeping the lights on 24/7 — just in case. And to justify, now we have various types of efficient lightbulbs on the market. We asked the question: how can we improve energy efficiency with the conventional method?
Led by Miro Salem, Global Head of AI and Autonomous Networks, Rakuten Symphony is orchestrating Synergy, an across-the-network energy orchestration platform. One of the components, invented by Krishna Kesavan, Chief Data Scientist, Synergy, deploys scalable, efficient and lightweight AI models that optimize server energy consumption at the granularity of the processor cores.
The team analyzed the total system architecture to reduce energy consumption by varying the speed of our processors. What initially suggested a possible 10% reduction of server energy cost now shows the ability to save over 30% with little to no negative impact on our network quality of service.
Today, testing in Japan, Synergy down cycles unnecessary processors from an end-to-end system perspective. Combined with the“lightbulb” approach, we achieve smarter network energy management.
Every step we take leads to a future with sustainable and smart energy. Working harmoniously with Rakuten Group and its partners, Rakuten Symphony not only looks to perfect energy management but to lessen its need.
We further our ambitions by creating more energy-efficient methods, especially hardware technology that reduces base energy profile. Our teams continue to explore new materials that can better reflect radio waves to cut back on base stations. In software, we are automation obsessed, creating Machine Learning (ML) models aiming at improving efficiency and optimization of our network and operations. Our operational engineers and various supporting mechanisms are also from an energy-conscious scope. Our air cites continued to grow through the years, while headcount remained consistent and one of the lowest in the industry. Our innovation, along with the comprehensive mentality for a better, smarter future, never desists.
What is important to us is delivering high-quality, reliable services while rectifying the telco industry's energy-conscious mindset. To learn more about Rakuten Symphony’s state of the art technology, visit here.