Using AI to slash 5G emissions and hit enterprise ESG goals

Implementing AI solutions in 5G networks can drastically slash energy emissions and accelerate progress toward critical ESG goals.

5G energy efficiency is often viewed as a line item for network operators, but new data suggests it is a macroeconomic lever capable of removing millions of tonnes of CO2 from national carbon footprints.

A study published in Resources, Conservation, and Recycling, conducted by the University of Surrey and Tsinghua University, counters the assumption that faster connectivity inevitably spikes energy demand. The research indicates that specific technical interventions – specifically, AI-driven sleep modes and intelligent hardware – can decouple data growth from emissions.

The ICT sector currently accounts for at least 1.7 percent of global greenhouse gas emissions. As operators densify networks to meet bandwidth targets, that figure risks climbing. However, the Surrey team found that optimising the network edge does not just lower the electricity bill for the operator; it reduces indirect emissions across the supply chain, particularly for heavy data users in finance and IT services.

Reducing 5G emissions with AI, sleep modes, and intelligent hardware

The most effective gains come from changing how the network behaves when idle. Operators have historically used binary sleep modes (i.e. turning components on or off.) The industry is now adopting more granular states: micro-sleep, light sleep, and deep sleep.

The study highlights AI-powered sleep control as a primary method for reducing emissions. By using deep reinforcement learning, base stations can predict traffic spikes and adjust power states dynamically rather than following a rigid schedule.

Hardware architecture is also evolving. The analysis identifies “cluster zooming” within cell-free massive MIMO (multiple-input multiple-output) networks as a high-performance strategy. This allows antennas to adjust their coverage area based on user density.

When combined with cell-free architecture (which reduces interference between access points) cluster zooming achieved energy efficiency ratings of approximately 91 percent compared to baseline operations.

Reconfigurable Intelligent Surfaces (RIS) also appear in the data as a viable efficiency tool to lower the emissions from 5G networks. These smart panels redirect radio waves with minimal power, improving signal propagation without requiring additional active transmitters.

Network efficiency extends to the user terminal. The aggregate energy consumption of mobile devices is a massive and often uncounted variable in the ICT footprint. The research suggests that refined signalling protocols on handsets can deliver substantial economy-wide reductions.

Techniques such as dynamic indication of control channel monitoring allow devices to check for signals less frequently when not in active use. For enterprises managing large fleets of mobile devices or IoT sensors, these protocols extend hardware lifecycles and reduce charging frequency.

Addressing supply chain visibility

To quantify these impacts, the researchers used an environmentally extended input-output (EEIO) model. Unlike standard Life Cycle Assessments that look at a single product, this model traces impacts across 33 sectors of the UK economy.

Dr Lirong Liu, Associate Professor at Surrey’s Centre for Environment and Sustainability, explained that the model exposes the hidden carbon cost of data. “Smarter base stations and devices don’t just cut electricity use in telecoms—they reduce indirect emissions in the whole supply chain,” Liu said.

“The modelling framework allowed us to quantify effects that are usually hidden, especially the indirect emissions linked to electricity use and wider supply chains. It also gave us a clear way to compare different 5G features side-by-side and identify which combinations deliver the strongest environmental benefits.”

The data shows that the financial, IT services, and programming sectors see the largest indirect benefits from these network upgrades. As these industries rely heavily on real-time data transmission, the carbon intensity of the 5G network becomes a component of their own Scope 3 emissions.

Regulatory levers to promote ESG goals

The technology to achieve these savings exists today. Professor Pei Xiao, from Surrey’s Institute for Communication Systems, noted that “many of these energy-efficient features are already on the engineering roadmap.”

“What this study provides is a clear system-level view of where the biggest carbon wins lie—and why regulators, operators, and industry should prioritise them as part of the UK’s net zero transition,” Xiao said.

Adoption may require policy intervention. The research suggests that 5G policy needs to move beyond simple coverage metrics. Future spectrum licenses could include energy targets, forcing operators to demonstrate network efficiency to access 5G frequencies.

For wholesale carriers and TowerCos, this regulatory shift would likely flow down into service level agreements. Enterprise clients, facing their own net-zero mandates, may soon require operators to prove they are actively lowering the emissions from their 5G services by utilising these low-power architectures.

See also: Samsung: Turning legacy infrastructure into AI-ready networks

Want to learn more about AI and big data from industry leaders? Check out AI & Big Data Expo taking place in Amsterdam, California, and London. The comprehensive event is part of TechEx and is co-located with other leading technology events including the IoT Tech Expo and Cyber Security & Cloud Expo. Click here for more information.

Telecoms is powered by TechForge Media. Explore other upcoming enterprise technology events and webinars here.