Ericsson rolls out AI in RAN subscription

Ericsson has launched a new software subscription model designed to integrate telecommunications-grade artificial intelligence directly into radio access network infrastructure. The Swedish vendor intends for this service to allow mobile network operators to deploy advanced machine learning models within existing baseband and radio units. By focusing on software-based enhancements, the company aims to improve 5G performance, automation, and total energy efficiency without requiring significant investments in additional hardware.

The subscription service, known as AI in RAN, targets the growing complexity of modern cellular networks by providing automated tools for real-time adjustments. Ericsson maintains that the integration of these AI models will assist operators in managing the increasing volume of data traffic while simultaneously reducing the operational expenditure associated with manual network tuning. The software is designed to operate seamlessly across different generations of hardware already deployed in the field.

A primary objective of this new offering is the optimisation of spectrum usage and power consumption across the radio access network. The AI algorithms are capable of predicting traffic patterns and adjusting transmission power levels in millisecond intervals, which leads to notable reductions in electricity usage during periods of low activity. This approach aligns with broader industry trends where service providers are seeking sustainable methods to scale their infrastructure without increasing their carbon footprint.

The software-centric approach also addresses the need for greater network agility as 5G transitions toward more advanced iterations. By utilising a subscription model, Ericsson enables operators to access continuous updates and improvements to the AI models as new network challenges arise. This ensures that the radio access network remains capable of supporting low-latency applications and massive machine-type communications as these services gain wider adoption in the enterprise sector.

Furthermore, the introduction of AI in RAN simplifies the path toward fully autonomous network operations. The system reduces the reliance on human intervention for routine tasks such as interference management and load balancing, which allows technical teams to focus on higher-level strategic planning. This shift is expected to enhance the overall reliability of 5G services by identifying and mitigating potential performance bottlenecks before they impact the end-user experience.

Ericsson has confirmed that the new software capabilities are compatible with its latest generation of RAN products and can be activated via standard remote updates. The subscription package includes ongoing support and analytical tools to help operators measure the specific impact of the AI models on their unique network configurations. This ensures that the deployment provides measurable value in terms of both quality of service and cost savings.

The strategic move represents a pivot toward software-led revenue streams for the equipment manufacturer as the global pace of new 5G hardware deployments shows signs of stabilisation. By offering a subscription service, Ericsson provides a more flexible commercial framework for operators who may be constrained by capital expenditure budgets. This model allows for a more gradual implementation of advanced features based on specific regional demands and traffic requirements.

Operators currently testing the new software capabilities are expected to report on initial performance gains and energy savings throughout the coming fiscal quarters. As more service providers integrate these AI-driven features, the industry will likely see a shift in how radio resources are managed and optimised at the edge of the network. The global rollout of the subscription service marks a significant step in the evolution of intelligent, self-optimising telecommunications infrastructure.