How AI is reshaping RAN: the real benefits you need to know

According to McKinsey, telcos spend 60% to 70% of budgets on operations and services. Wouldn’t it be good to bring that down and drive greater profitability? Now you can: deploying AI in the RAN can help mobile network operators (MNOs) drive major savings while enhancing customer experience (CX).

RANs remain the backbone of wireless communications, and they have always needed human workforces to design, deploy, and optimize them. However, as demand from end-users for faster speeds, lower latency, and greater reliability grows as 5G increases, network management becomes more complex, time-consuming, and costly. AI can help.

AI enabling new RAN efficiencies

MNOs are always looking for places to cut costs. And that’s understandable: it’s a highly competitive market in which average revenue per user (ARPU) has fallen by around 20% over the past ten years. AI can give MNOs the power to cut costs across all kinds of areas of the business, using automation, self-organizing networks, predictive maintenance, load balancing, and more.

Optimizing RAN performance is always a challenge for MNOs and one that consumes resources. End-users expect optimum coverage, capacity, and quality of service, but that requires constant adjusting and fine-tuning of the network. As demand for data continues to increase exponentially, analyzing and exploiting it becomes impractical.

AI can improve self-organizing networks (SON), increasing automation of configuration, optimization, and troubleshooting and making the network more responsive to real-time changes. The net result is enhanced performance across metrics like latency, throughput, and reliability. Some MNOs are already benefiting: Vodafone’s Zero Touch Operation Strategy deploys AI in RAN and targets 50% fault prevention , while Nokia’s AI/ML-powered SON detects RAN issues 120x faster. Analysys Mason reports that 89% of CSPs believe AI in RAN will reduce TCO , and around two-thirds expect reduced downtime from more self-healing and predictive networks plus improved spectrum efficiency.

Energy is another area where AI can help MNOs drive savings. The GSMA reports that MNOs spend 20% of OPEX on energy and Ericsson's Intelligent Radio Access Network (RAN) Power Saving is an example of an AI solution that can deliver 20% power savings across 5G daily operations.

AI-RAN: big impact on CX

AI can empower MNOs to improve CX by analyzing user behavior and preferences, enabling them to offer more personalized services. It will allow CSPs to predict when and where users need the most bandwidth for heavy-usage apps like video streaming or gaming and automatically adjust network resources accordingly, again setting the stage for more personalized services tailored to end-users’ activities.

MNOs can deliver improved CX through proactive and predictive network management, using AI to monitor real-time network conditions and forecast potential slowdowns or connectivity issues. They can then quickly take corrective actions like reallocating bandwidth or rerouting traffic. Predictive maintenance is another big bonus that AI-RAN delivers to CX, being used to analyze historical network data and identify patterns, and predict possible equipment failures. It’s a step along the road to self-healing networks, which automatically detect and rectify faults without human intervention. It all adds up: 70% of customers say CX is a significant factor when picking their MNO, and more loyal customers also tend to spend more ARPU.

AI in the evolution of 5G antennas

In the context of AI-driven RAN advancements, AI is transforming the future of wireless communication, especially in the evolution of 5G and beyond. By integrating artificial intelligence into 3D multiple input and multiple output (MIMO) base station antennas (BSA), the technology can meet the growing demands of users more efficiently. AI helps optimize how these antennas handle signals, leading to faster data speeds and better connectivity. With AI, these antennas can adjust and improve their performance in real-time, ensuring a reliable and high-quality experience, whether it's for video calls, streaming, or handling massive data transfers.

Additionally, AI-driven optimization significantly enhances energy efficiency by minimizing unnecessary power usage, reducing operational costs for CSPs, and supporting environmental sustainability. By intelligently managing power consumption, AI-enabled antennas contribute to a greener, more energy-conscious infrastructure, helping lower wireless networks' carbon footprint. This combination of AI and advanced antenna technology is key to delivering better energy efficiency and greater performance in our connected world.

Furthermore, AI is increasingly transforming antenna systems in RAN by utilizing advanced techniques like machine learning for beamforming and adaptive tuning. With AI-driven algorithms, antennas can dynamically adjust parameters such as tilt, fine-tune azimuth beamwidth, and optimize transmit power in real-time, allowing for precise adaptation to traffic patterns, maximizing coverage and performance, and minimizing interference. For antenna vendors, this opens up exciting opportunities to provide smarter, more responsive antenna systems that align with the future demands of CSPs. AI-powered antennas can be especially impactful, particularly in regions such as APAC and MEA, where there is rapid subscriber growth, diverse geographical challenges, and significant urban-rural disparity. They can help manage complex topographies, provide more consistent connectivity in densely populated urban centers, and extend reliable coverage to underserved rural areas.

By integrating AI into antenna solutions, vendors can deliver products that help CSPs improve network coverage, boost service quality, and minimize churn - all while reducing the need for costly manual optimization. This AI-driven efficiency makes it possible for CSPs to handle increasing data loads and support new applications with a scalable, energy-efficient, future-ready infrastructure solution, making advanced antenna systems a critical part of next-generation network evolution.

Possibilities and potential for the future

AI is going to have a major impact on the future of the RAN. By enabling intelligent network optimization, proactive CX enhancements, and more efficient resource management, AI can transform how MNOs manage the network. According to a recent report from Dell’Oro Group, 6G RAN revenues are projected to reach nearly $30 billion by 2033, with Sub-7 GHz and cmWave macros anticipated to play a dominant role in the 6G landscape. It all points to some exciting and transformative times.

The industry certainly seems to agree. In February this year, the AI-RAN Alliance was launched by some of the biggest players in telecoms, including Ericsson Networks , 英伟达 , 诺基亚 , and T-Mobile . The Alliance aims to collaborate on innovation projects that can maximize AI’s potential in RAN, and working groups established so far include the AI-for-RAN Working Group, to explore and enhance the use of AI in RANs to improve performance metrics, such as efficiency and capacity. There is the AI-and-RAN Working Group, with a mission to explore the use of converged computer-and-communications infrastructure to run RAN, AI, and GenAI workloads, enhancing platform utilization and creating new monetization opportunities.

The third is the AI-on-RAN Working Group, which is tasked with defining radio interface requirements to run AI and GenAI applications across the consumer, enterprise, and government sectors. This working group’s goal is to benchmark the performance of these applications on 5G and identify new requirements that will support future 6G systems.

I look forward to seeing what comes from it all and how AI will continue to support the evolution of RAN and antennas. AI in the global telecoms industry is predicted to grow from a market revenue of $2.2 billion in 2023 to a massive $19.5 billion in 2030 .