Green Telecom Networks: Artificial Intelligence Solutions for Improving Energy Efficiency

It is no secret that green energy and reducing energy consumption are the future. A growing number of operators have taken a leading role in sustainability and the use of renewables to meet or exceed these decarbonization goals, and these will expand in the 5G era. As the world works to find cleaner sources of power, Green Telecom Networks (GTN) are also coming under scrutiny as an area where green solutions can be applied.

Future telecommunications networks need to be sustainable, and in recent years there have been many advances in this area thanks to Artificial Intelligence (AI) solutions. This article will highlight the most critical solutions based on AI technologies that enhance network energy efficiency.

World’s Climate

Working towards a sustainable future is no longer an option. At the same time, there is no apparent force to build a solid fundamental mindset for change. The world is waiting for a plan to connect the global economy that incorporates sustainability with inclusion and growth.

• Growth - without growth; there is no way to achieve prosperity and wellbeing or pay for transitions needed for a sustainable economy;
• Inclusion - to ensure demand for growth, inclusion has to take place, it must be connected with new ways of work thus opportunities for being more productive and firmly sealed with quality of life for all citizens;
• Sustainability - sustainable and inclusive growth can be a dynamic, self-reinforcing combination, but achieving it will require addressing counteracting forces.

Global Energy Consumption

Global energy demand increased by 4.6% in 2021, more than offsetting the 4% contraction in 2020 and pushing demand 0.5% above 2019 levels. This represents the 10th consecutive year the world set a new all-time high for energy consumption. Almost 70% of the projected increase in global energy demand is in emerging markets and developing economies, where demand is set to rise to 3.4% above 2019 levels. Energy use in advanced economies is on course to be 3% below pre-COVID-19 levels.

Although energy consumption worldwide and industries are significantly increasing, energy savings have become a top priority for almost every country and industry sector.

According to International Energy Agency (IEA) World Energy Outlook, renewable energy should account for 27 percent of global energy consumption in 2050 in the light of changing policies and declining costs.

Environmental Priorities in the Telecom Sector

Production of telecommunications network infrastructure equipment causes unneglectable Greenhouse Gas (GHG) emissions and consumes many fossil fuels. These include various metals and plastics. Scarce or problematic materials are also used, such as Critical Raw Materials (CRM), conflict minerals, and rare earth elements.

When equipment is made from primary sources (i.e., materials derived from mined ores or crude oil used for plastics), these materials are referred to as primary materials, and the supply chain starts from the mine and the oil wells, which both have to be considered as part of the life cycle impact of the product. In particular, as mine and oil production are associated with land use and pollution, they are both related to the effects on the landscape, biodiversity, and natural habitats.

Likewise, using secondary materials (i.e., recycled materials) instead of primary materials in products is a necessary eco-design action alongside designing recyclable products. Moreover, using recycled materials is also associated with lower GHG emissions; this effect could be substantial for materials relying on energy-intensive processes such as aluminum.

To work towards full circularity, equipment needs to use recycled materials, be recycled at the end of life, and move towards phasing out primary materials. Considering the use of materials, two main material fractions need to be considered; metals and plastics. While the recovery and reuse of some metals, such as aluminum, copper, steel, etc., is already relatively well established, the use of recycled plastic still needs to be expanded.

Circularity Aspect

The concept of circular economy, which is closely related to materials efficiency, has become very popular and needs further implementation. This concept includes multiple aspects associated with prolonging the lifetime of products and increasing the circularity of subsequent steps, including factors such as durability, repair, reuse, upgradability, remanufacturing, recyclability, use of recycled materials, and use of critical raw materials.

At the end of life, preferably after a second life of the product, recycling of materials is the main objective, and this recycling needs to be economically feasible to happen. Different effects need to be considered together, such as the use of energy, recyclability, use of recycled materials. In any case, the overall impact throughout the life cycle should be a guiding star to optimize the design.

According to existing legal regulations, which apply to many industries, including telecom network operators, equipment should have a recyclability rate as close to 100% as possible.

Environmental Footprint Reduction

Products associated with the telecom industry and data centers work in a 24/7 life span. This makes most greenhouse gas (GHG) emissions and several other impacts related to their usage. This implies a hierarchy of how to deal with this equipment to reduce their environmental impact and minimize the use of material:

Length of Use

The longer the equipment is used, the better the GHG emissions spread over the lifetime during production. It means that circuitry and mechanical parts are designed with this in mind. Therefore, certain design aspects need to be considered.

Second Life

If products are phased out before their technical life ends, a subsequent secondary use must be facilitated. In principle, the same design considerations apply for the primary use.

Parts Reuse

If the length of use and second life is exhausted due to, e.g., the equipment is worn out and/or beyond economic repair, identification, and reuse of any usable part is a preferred option.

Recycling

When either first, second, or third are exhausted, material recycling as close as possible to 100% needs to be possible.

Energy Efficiency for the Telecom Industry

For Wireless Networks, energy efficiency can be defined in many ways. First of all, it has to be determined how much energy is needed to deliver one unit of data, serve one connection to supply one base station, and produce one revenue unit.

Each measure has its advantages and disadvantages, so the choice always becomes a question of balance. Only taking into account the total amount of used energy would be unfair and biased as more data than ever before is being transferred across Mobile Networks.

While energy use can be a high cost, it is also an area with many opportunities for operating expenses (OPEX) savings if the required capital investments/upgrades are made to network equipment.

Recent studies from leading vendors and CSPs show that energy costs for the telecom industry worldwide are already high: at the end of 2018, they accounted - on average - for around 5 percent of operating expenditures. In emerging markets, where low grid coverage often means operators must supply their power with a generator set, energy can account for as much as 7 percent of expenditures. Costs look set to rise further, putting more significant pressure on margins at a time when the industry can scarcely handle any additional financial burden.

This growing energy challenge results from the exponential growth in traffic that new 5G services are likely to deliver. Although the 5G NR (New Radio) standard is more energy-efficient per gigabyte than the 4G standards, the proposed 5G use cases and new spectrum bands will require many more mobile sites and Base Stations (BTS’s), outstripping potential energy efficiencies. According to industry estimates, each 5G site will need two to three times more power than the 4G-equivalent site. At the same time, as more services are provided at the Edge, the number of data centers will need to rise. These circumstances already account for 5 to 10 percent of a telecom operator’s energy costs.

Costs are not the only concern. According to GSM Association, the telecommunications sector already account for 2 to 3 percent of total global energy demand. As operators’ energy consumption expands, so will their carbon footprint, hurting the environment and their reputation and standing, particularly among the expanding class of socially responsible investors.

But this does not have to be the case. All operators have considerable scope to cut energy costs and consumption. For instance, transferring data consumes only around 15 percent of energy in one selected wireless network. The rest of 85 percent is wasted because of heat loss in power amplifiers, equipment kept idling when there is no data transmission, and inefficiency in rectifiers, cooling systems, and battery units.

Some savings lie in deploying Artificial Intelligence (AI) and the Internet of Things (IoT): some in structural and architectural transformations, and some in cheaper and more sustainable energy sourcing. The extent of the potential savings will vary by operator and market. Regulations governing distribution and retailing, green-energy incentives, OEM choice, and an operator’s starting point in energy efficiency will all make a difference. Nevertheless, studies confirm that many operators can reduce energy consumption and costs by at least 15 to 20 percent in the space of just one year—and moreover a more extended period.

Telco’s Views on Sustainability

According to the latest research, at the end of 2021, the telecoms industry served more than 5.3 billion unique mobile subscribers worldwide, equivalent to 70% of the population. The mobile sector has worked collaboratively to create an industry-wide climate action roadmap to achieve net-zero greenhouse gas (GHG) emissions by 2050, according to the Paris Agreement. Over 30% of carriers have made public commitments.

Energy Cost Reduction Opportunities

The enormous opportunities for energy consumption and cost reductions lie in many areas:

Structural and Architectural

Energy is the primary source of cost savings when decommissioning legacy networks; for example, moving a stand-alone 2G network to a Single RAN (SRAN) architecture is more energy-intensive. Similarly, migrating to architectures such as Cloud RAN can deliver energy savings of more than 10 percent.

Another opportunity lies in customizing RAN and passive infrastructure specifications site by site, depending on load conditions. One size does not need to fit all. The size of battery trays for backup power can be custom designed, for example, as can multiple-input and multiple-output (MIMO) configurations.

Internet of Things (IoT) Enabled Energy Optimization

It is hard to reduce energy consumption and costs if consumption measurements are not done accurately, but until recently, accurate measurement on an industrial scale has been difficult. Old equipment does not always measure consumption, and even if it does, recording it requires hundreds of employees to make and submit accurate readings.

The IoT uses sensors to track consumption. That advance opens up all sorts of new ways for operators to save energy: i.e., sensors that read consumption, intelligent meters, give companies access to the time of usage discounts in markets. Significantly, companies can retrofit the IoT for use by old equipment.

Sustainable Energy Sourcing

Operators often fail to give enough attention to the supply sides of their energy expenditures. It is still rare to find an energy sourcing specialist in a telecom operator’s procurement department, even though sizeable cost savings lie in better sourcing.

Some areas like procurement can directly purchase energy from power suppliers at a better rate. Green energy suppliers are cheaper in some markets; thus, telecom operators should take advantage of price differentials. Also, there are opportunities for companies to cut costs by generating their green power on-site in some cases.

Artificial Intelligence Technology

Running systems that are not in constant use consumes significant amounts of energy. Typically, Radio Access Network (RAN) accounts for about 60% of the power used at a mobile site. Data traffic loads are intermittent, though, so different RAN parts can be put briefly into sleep mode, even during periods of peak traffic. In such cases, energy-conserving AI tools can deliver 5 to 7 percent savings for some operators, in addition to protection that accrues from stand-alone, site-level efficiency measures.

Similar energy-saving advances are occurring with AI on Fixed Networks. For example, AI can reduce the energy cost of central offices by between 3 and 5 percent by continuously calibrating the optimal settings of chillers, pumps, and fans to guard against waste.

Transformation Plan

Although there is enormous potential to reduce the total energy consumption and costs of the telecommunications network, realizing that potential is not an easy task.

Usually, taking up such an essential topic as sustainable development requires involvement on many levels in the organization and mindset changes. Energy costs may well be aggregated at the company level. Still, responsibility for the cost drivers is split across many different roles and business units, including network and infrastructure planning, field operations, facilities management, procurement, IT.

Moreover, reducing energy costs requires resources (labor and capital) when operators are concentrating their investments on expanding the capacity and reach of their networks. For both these reasons, energy consumption and costs are unlikely to fall without high-level recognition of the importance of reducing them and a comprehensive strategy for doing so.

Wireless networks operators have many right questions to ask themselves. These questions, in turn, should lead to the correct answers, which are solutions that meet the needs of the times. Ultimately, these answers should be brought down to a proper plan, which is a green strategy.

So what constitutes the alleged green strategy? Is it:

1. Embedded sustainability?
2. Inclusive growth?
3. Engaged ecosystem? , or maybe the telco sector needs to go back to basics and
4. Build the awareness of employees, customers, plus end users about reducing energy consumption, energy efficiency, and ways for renewable energy?

Drivers Behind the Urgency

Recent studies highlight enormous pressure to be expected on mobile networks:

Increasing Demand for Connectivity

5G upgrades will accelerate post-pandemic. By 2025, 5G mobile connections will reach around 2 billion, or 20% of the global customer base. These subscribers are likely to have monthly data consumption rates higher than LTE. The growing number of connections and data traffic is expected to increase energy demand.

More Complex Networks

Early readings suggest that the total power consumption of a 5G base station is significantly higher than that of a 4G base station. Although 5G is more energy-efficient than previous wireless technologies, absolute energy consumption is more significant because of higher traffic demand and complex new 5G use cases. With carbon emissions being a top concern for industry, regulators, investors, and customers, there is ample motivation to reduce consumption and shift to renewables and alternative energy sources.

Massive 5G Connection by the End of 2025

Despite the impact of the COVID-19 pandemic, it is forecasted that 5G rollouts on the global market will generate 2.2 billion new 5G connections.

Artificial Intelligence and Machine Learning

New technologies such as Artificial intelligence (AI) have emerged as an essential energy-efficiency enabler for telcos, according to a new study conducted by GSMA Intelligence.

Ways to Build a Green Network

AI in Network Design

Network design is the foundation of energy-efficient networks. Operators are deploying a wide range of new cells during the start of the 5G era, including mmWave cells and small cells integrated into street furniture, street lamps, or even indoor sites. Propagation analysis and site selection is a complex and labor-intensive task.

AI can help here, not just speed up the process but also make it more accurate. An AI algorithm first creates a propagation map; then, after some additional input from the operator, an optimal network configuration can be reached. Initial on-site performance checks can feed information back into the algorithm to validate and finetune the network planning.

No Physical Tests

Operators have historically spent significant resources to test their service quality. Constant testing and mapping of users’ experience under different conditions at different times of the day is essential to provide high-quality telecoms services.

With the help of AI, operators can reduce the size of testing crews, their travel time, and the related petrol and electricity costs. Network tuning and virtual test drives can rely on real-time traffic data, as every user can become a test drive, and AI can help to form actionable insights for the operators and reduce the need for physical test drives.

Data-centric Energy Management

In 2021, AI and ML-driven applications primarily focused on shutdown solutions in the RAN, but energy management should be addressed in an end-to-end manner. Customers expect to have an overall understanding of all their data and to use holistic solutions. To achieve this, the complete digitization of the network infrastructure is needed, with sensors and site controllers on every network element. Most operators are currently working to achieve this holistic view and build their data pipelines, including active and passive infrastructures.

Passive Infrastructure

Depending on the climate and the quality of the electricity grid, passive infrastructure, especially air conditioning, is responsible for a significant part of operators’ energy use. Network equipment does not usually measure energy consumption and many aspects of the passive infrastructure lack metering.

Most operators currently have limited information on their passive infrastructures, such as the efficiency of their diesel generators and rectifiers. Even if the equipment has the metering capacity, recording the data would be labor-intensive and not in real-time. After operators deploy the required sensors and build the data pipelines, they will be able to view the energy requirements and usage from passive infrastructure and increase the impact of AI.

Deep Dive Into AI-Driven Energy Efficiency Improvements

Artificial Intelligence (AI) and Machine Learning (ML) are delivering new capabilities and efficiency improvements to the telecommunications sector in more ways than ever before. They allow network equipment to perceive, reason, insights and provide new methods for solving technical challenges.

Energy management is essential data driven, and operators cannot efficiently process information and make accurate time decisions at scale without the use of AI. Holistic and end-to-end AI and ML can provide an omnipresent system-level approach that improves energy efficiency across hardware, software, and algorithms. This becomes particularly important with 5G networks, given that average customer data usage on 5G is structurally higher (5–10×) compared to LTE, which puts heavy pressure on energy consumption.

Green Mobile Networks Solutions

Telefónica has tested a new liquid cooling system that, according to the operator, can increase the energy-saving of their facilities such as central offices and data centers by up to 50%. The system looks like a giant scanner, but it is an immersion cooling unit that Telefónica has put through its paces at its central office in the Spanish capital of Madrid as part of its efforts to identify new systems that can help it reach its carbon footprint reduction goals.

The operator says it has widely deployed the method of ‘free cooling’ at its various buildings and facilities, including BTS’s, whereby air from outside the facilities is circulated to help reduce energy costs, excellent equipment, and help to reduce energy consumption by up to 30%. When a system comes into play, it can increase energy efficiency by up to 50%, reduce carbon emissions from the use of refrigerant gases, reduce the footprint and improve the total cost of ownership of data centers.

Orange is trying to improve energy efficiency in their networks through the Green IT & Networks plan, which aims to optimize their technical deployments, increase the eco-efficiency of data centers, and cut the energy consumed by their routers. They are replacing old infrastructure with modern and eco-efficient equipment using alternative solutions such as passive cooling or solar power. The program has led to significant reductions in the Group’s energy consumption and associated carbon emissions since 2010.

South Korean SK Telecom shows innovative products and services promise greenhouse gas emissions reduction. Under the theme of ‘SK’s 2030 Net Zero Pledge’, the company will showcase a plethora of green technologies, with a primary goal of countering the common perception that AI and telecoms network deployments are a significant contributor to environmental pollution of the power consumption required for their operation. To help overturn that perception, SKT will put the spotlight on SAPEON, an Artificial Intelligence (AI) chip it developed to accelerate tasks and process vast amounts of data more efficiently, thus requiring less power usage.

In a new study, Nokia and Telefónica together have determined that 5G networks are up to 90 percent more energy efficient than legacy 4G networks. For three months, Nokia worked with Telefónica on a joint research project that looked in granular detail at the Spanish telco’s radio access network (RAN) power consumption. Eleven different traffic load scenarios were measured to determine the energy consumed per Mbps on the load distribution and showed, with absolute certainty, that 5G RAN technology is much more efficient than 4G in terms of energy consumption thanks to a raft of new hardware and software features at both the base stations (BTS’s) and network levels. The research, which used Nokia’s AirScale portfolio of 5G equipment, shows that 5G carries more bits of data per kilowatt of energy than any previous generation of wireless technology and is, by its very nature, much greener than its predecessors. Of course, 5G is a work in progress, and the new networks will require further tuning and refinements to further enhance energy efficiency and minimize carbon emissions.

All the electricity Vodafone uses to power its operations in Europe has come from renewable energy sources. It does mean its entire European operations, too: mobile and fixed networks, data centers, retail outlets, and offices are all included. To add further context, Vodafone noted that 96% of its total energy usage in the 2021 financial year to the end of March came in the form of purchased electricity from the different utility providers. As a result of its efforts, all of its purchased electricity in Europe – including cases in which Vodafone is a tenant on other landlords’ sites – will now come from renewable generation sources such as wind farms, solar panels, and hydro.

Green Telecom Networks Outlook

Much effort and discussions on many levels still need to occur for mobile operators, customers, and vendors to understand that challenges and solutions go hand in hand. There will be no Green Networks without a common front and knowledge exchange.

Currently, Europe is at the forefront of Green Telecom Networks. European operators are committed to leading the global transition towards a zero-carbon economy, including supporting national climate change policies and the European Green Deal.

The CEOs of 13 telecoms firms were among the 26 initial signatories of the European Green Digital Coalition – a cross-sector agreement to take action to support the green and digital transformation of the EU. The chief executives committed on behalf of their companies to take action in the following areas:

• Investing in the development and deployment of greener digital technologies and services that are more energy - and material-efficient;
• Developing methods and tools to measure the net impact of green digital technologies on the environment and climate by joining forces with NGOs and relevant expert organizations;
• Co-creating, with representatives of other sectors, recommendations and guidelines for the green digital transformation of these sectors that benefit the environment, society, and economy;

Closing Remarks

As worldwide mobile networks infrastructure grows and the power consumed by the telco industry field increases, the importance of Green Networking solutions is crucial for sustainable growth. To continue advancing in the network’s environmental footprint reduction, there should be increased efforts to develop standard methodologies to solve the complex question of assessing environmental footprint services.

Ways to Overcome Challenges

1. Work across the industry to develop new standards, harmonization, and processes to scale up solutions,
2. Enable AI-driven network management functions with a pool of global experts,
3. Offer flexibility and AI as a service (AIaaS) or energy-saving as a service (ESaaS),
4. Offer a high level of standards, local experts, and knowledge of the local environment.