Implementing AI-driven solutions within 5G networks can significantly reduce energy-related emissions and accelerate progress toward corporate and national ESG targets.
Energy efficiency in 5G networks is often treated as an operational cost for network operators. However, new research shows it can function as a macroeconomic lever, potentially removing millions of tonnes of CO2 from national carbon inventories.
A study published in Resources, Conservation, and Recycling, conducted by the University of Surrey and Tsinghua University, challenges the assumption that faster connectivity necessarily increases energy consumption. The research finds that targeted technical interventions—especially AI-managed sleep modes and smarter hardware—can decouple data growth from emissions.
The ICT sector currently contributes at least 1.7 percent of global greenhouse gas emissions. As operators build denser networks to meet rising bandwidth demand, that share could increase. The Surrey–Tsinghua analysis shows that optimizing the network edge not only cuts operators’ electricity costs but also reduces indirect emissions across supply chains, particularly for high-data industries such as finance and IT services.
Reducing 5G emissions with AI, sleep modes, and intelligent hardware
The largest efficiency gains come from changing how networks behave when idle. Historically, operators used binary sleep modes—components are either fully on or off. The industry is shifting toward more granular power states: micro-sleep, light sleep, and deep sleep.
The study highlights AI-powered sleep control as a primary method to cut emissions. Using deep reinforcement learning, base stations can forecast traffic patterns and shift power states dynamically, rather than relying on fixed schedules.
Hardware architectures are also evolving. The research identifies “cluster zooming” within cell-free massive MIMO (multiple-input multiple-output) deployments as a high-impact strategy. Cluster zooming lets antenna groups adapt their coverage footprint to user density.
When combined with cell-free architectures—which reduce interference among access points—cluster zooming produced energy-efficiency improvements of around 91 percent compared to baseline operations in the study’s simulations.
Reconfigurable Intelligent Surfaces (RIS) also appear in the findings as an effective low-power tool to improve signal propagation. These passive or semi-passive panels redirect radio waves with minimal energy use, enhancing coverage without adding active transmitters.
Network efficiency gains extend to user devices. The cumulative energy consumption of mobile handsets and IoT devices is a significant but often overlooked component of the ICT footprint. The research shows that better signalling protocols on devices can yield meaningful, economy-wide energy reductions.
Techniques such as dynamically indicating when a device should monitor control channels let handsets check for network activity less frequently when idle. For enterprises managing large device fleets or sensor networks, these protocols extend hardware lifespans and reduce charging cycles.
Addressing supply chain visibility
To quantify these effects, the researchers used an environmentally extended input-output (EEIO) model. Unlike a single-product Life Cycle Assessment, this approach maps impacts across 33 sectors of the UK economy, revealing indirect emissions that standard methods can miss.
Dr Lirong Liu, Associate Professor at Surrey’s Centre for Environment and Sustainability, explained that this modelling exposes the hidden carbon cost of data. “Smarter base stations and devices don’t just cut electricity use in telecoms—they reduce indirect emissions across the supply chain,” Liu said.
“The modelling framework let us quantify impacts that usually remain hidden—especially indirect emissions tied to electricity use and broader supply chains. It also enabled side-by-side comparisons of different 5G features to identify which combinations deliver the largest environmental benefits.”
The study shows that the financial, IT services, and software development sectors stand to gain the most from these network upgrades. Because these industries rely heavily on real-time data transfer, the carbon intensity of underlying 5G services becomes a factor in their own Scope 3 emissions.
Regulatory levers to promote ESG goals
The technologies needed to realize these savings are available today. Professor Pei Xiao, from Surrey’s Institute for Communication Systems, noted that many of these energy-efficient features are already on engineering roadmaps.
“This study provides a system-level view of where the biggest carbon reductions are achievable—and why regulators, operators, and industry should prioritise those measures as part of the UK’s net-zero transition,” Xiao said.
Wider adoption may require policy support. The researchers suggest moving 5G policy beyond simple coverage metrics. Future spectrum licences could include energy-efficiency conditions, requiring operators to demonstrate measurable improvements before gaining access to frequencies.
For wholesale carriers and TowerCos, such regulatory changes would likely cascade into service-level agreements. Enterprise customers, working toward their own net-zero commitments, may increasingly demand proof that operators are reducing emissions by using low-power network architectures.
See also: Samsung: Turning legacy infrastructure into AI-ready networks
Want to learn more about AI and big data from industry leaders? Attend an AI & Big Data Expo event to hear from practitioners and vendors about real-world deployments and lessons learned. These events bring together experts across AI, IoT, cybersecurity, and cloud technologies and provide practical guidance for integrating energy-efficient architectures into enterprise networks.
Telecoms coverage is published by TechForge Media. Explore other enterprise technology events and webinars organised by the publisher for updates on sustainable networking and related topics.