By Catherine Sbeglia Nin December 2, 2024
Collected at: https://www.rcrwireless.com/20241202/fundamentals/traffic-energy-ai
Energy consumption is one of the biggest factors of OpEx, says AI-RAN Alliance Chair Alex Jinsung Choi
The reduction of operation expenses (OpEx) is a key priority for telecom operators, and according to Alex Jinsung Choi, chair of the AI-RAN Alliance and principal fellow of SoftBank Corp.’s Research Institute of Advanced Technology, energy consumption in the radio access network (RAN) is one of the biggest factors of OpEx.
“AI will play a big role in making RAN operations more sustainable,” he told RCR Wireless News. Specifically, AI and machine learning (ML) can identify lower network load and only scaling up resources when necessary, enabling telcos to fine-tune energy usage based on real-time demand. “This kind of dynamic scaling can help ensure that network performance targets are always met while significantly cutting down unnecessary energy usage,” said Choi. “It’s not just about using less energy; it’s about using the right amount of energy at the right time.”
One such example is smart radio frequency (RF) channel management in which Data Quality Monitoring (DQM) — a kind of reinforcement learning model — is leveraged to provide dynamic RF channel reconfiguration, leading to material energy savings. “Dynamically, we can change channels, cells and even signal levels on and off based on network traffic,” explained Choi, adding that this approach involves offloading users to neighboring cells before switching off certain cells to ensure seamless connectivity.
Nokia takes energy efficiency to the ‘extreme’
Nokia’s approach to energy efficiency has been self-described as “extreme.” For instance, the vendor takes this idea of spinning network resources up and down to reflect real-world traffic to the max with its Deep Sleep cell switch-off mode. This solution, announced in February, leverages software in the vendor’s AirScale Habrok Massive MIMO radio units to reduce energy consumption by up to 97% compared to a cell on air but without traffic. The company also has a ‘Zero traffic, zero energy’ solution that switches off all radio resources when there is zero traffic.
Additionally, ReefShark System-on-Chip (SoC), Nokia’s SoC technology dynamically adjusts internal resources based on traffic demands, contributing to a 15% reduction in energy consumption. Energy efficiency of the radio network software solutions can be further enhanced with the vendor’s MantaRay Energy solution, which automates and optimizes the configuration of the energy-saving software features.
It was this last solution that Ed highlighted when speaking with RCR Wireless News. The Nokia MantaRay SON solution, he shared, taps AI and ML to autonomously identify and address network inefficiencies by automatically switching off underutilized cell sites. “This can optimize the RAN energy based on the real traffic pattern,” he said, adding that Nokia also uses digital planning services like digital twins to further improve efficiencies. “It’s all [to] improve the energy efficiency for the entire site overall.”
In July, stc Group deployed Nokia’s MantaRay Cognitive SON in its commercial network. According to Nokia, stc’s network successfully maintained consistent connectivity despite traffic increasing by 40% during the measured period. Autonomous RAN operations also reduced manual work and improved network quality, the vendor claimed.
Vodafone and Chunghwa Telecom, among others, are also using the Noka MantaRay solution to achieve a more dynamic and energy efficient RAN.
Leave a Reply