en.Wedoany.com Reported - Ericsson has launched a solution called "AI in RAN," which integrates artificial intelligence models directly into the Radio Access Network (RAN), aiming to help operators manage increasingly dynamic networks without significantly increasing costs or energy consumption. As AI drives a surge in data traffic and autonomous applications, operators face growing pressure to meet real-time response demands on communication networks.

AI in RAN is a software subscription service that enables AI models to run in real time on wireless devices and baseband units without requiring additional hardware. Its goal is to empower networks with autonomous decision-making capabilities through millisecond-level resource optimization and dynamic responses. Ericsson stated that the first features will be launched in the second quarter of this year, with continuous integration improvements throughout 2026. In tests, the technology achieved up to a 20% improvement in downlink performance, a 10% increase in spectral efficiency, support for twice the number of users in high-demand scenarios, and coverage accuracy ranging from 90% to 95%.

Ericsson explained that AI in RAN includes AI models specifically trained for telecommunications environments, capable of continuous learning based on network data, and supports an AI agent architecture for higher autonomous decision-making. This solution is part of Ericsson's AI RAN vision, covering three layers: "AI for the network," "AI in the network," and "executing AI workloads on RAN infrastructure." The company believes this integration helps build programmable networks that can sense services and automatically adapt to application demands. The new commercial solution strengthens Ericsson's roadmap toward AI-native networks, which the industry considers one of the pillars of 6G.

This article is compiled by Wedoany. All AI citations must indicate the source as "Wedoany". If there is any infringement or other issues, please notify us promptly, and we will modify or delete it accordingly. Email: news@wedoany.com