New Delhi, June 11(CXO Media): Ericsson has introduced a new software offering that embeds artificial intelligence directly into radio access networks (RAN), as telecom operators look for ways to improve network performance, automate operations, and manage rising data traffic driven by AI-powered services.
Called AI in RAN, the software is designed to run AI models inside basebands and radios without requiring additional hardware upgrades. The launch comes as telecom companies worldwide explore AI-RAN technologies to improve network efficiency and prepare for future AI-native network architectures.
AI Moves Closer To The Network Edge
The software uses telco-specific AI models that operate in real time within the radio network. Ericsson said the platform supports continuous learning through network data and can also enable agentic AI capabilities for advanced automation and network management.
The company plans to make the first AI in RAN features available during the second quarter of 2026. These include AI-native scheduling, AI-powered positioning, beamforming optimization, multi-layer coordination, and enhanced network observability tools.
Focus On Performance And Spectrum Efficiency
According to Ericsson, trials and commercial deployments across more than 15 networks have shown improvements of up to 20% in downlink throughput and up to 10% in spectral efficiency. The company also reported gains in user-positioning accuracy and support for a larger number of high-traffic users.
Several telecom operators, including SoftBank, Bell, SK Telecom, and Rogers, said they are evaluating or deploying AI-driven RAN capabilities to improve network performance, energy efficiency, and operational automation.
The announcement reflects a broader industry trend toward integrating artificial intelligence into core telecom infrastructure as operators seek to extract more capacity and efficiency from existing 5G investments while preparing for future 6G-era networks.
Also Read: TCS To Train 50,000 Employees As Anthropic Deal Shifts AI From Pilot Projects To Enterprise Use
