Ericsson launches AI in RAN software to boost 5G video performance
Ericsson has launched AI in RAN, a software subscription integrating AI models into 5G basebands and radios to enhance network performance, efficiency, and automation. This offering aims to provide immediate benefits for 5G networks, supporting a shift to AI-native RAN without requiring additional hardware and delivering up to 20% higher downlink throughput. The first features are available in Q2 2026, with further enhancements later in the year, and have been proven in over 15 deployments worldwide.
Key Takeaways
- Software-only upgrade delivers up to 20% higher downlink throughput and 10% better spectral efficiency.
- Integrated AI models perform ultra-low latency inference at the microsecond level for real-time radio optimization.
- Proven in commercial trials with SoftBank, Bell, Rogers, and SK Telecom prior to its Q2 2026 launch.
- New features include an AI-native Scheduler for Link Adaptation and AI-managed Beamforming to handle high-traffic users.
Why It Matters
For streaming providers, this shift brings deterministic performance to the network edge, potentially reducing buffering and improving 4K/8K delivery in congested 5G cells. By moving radio management from rigid rule-based logic to predictive AI models, Ericsson allows operators to extract more capacity from existing spectrum, which is critical as high-bitrate AR and video services stress uplink and downlink channels. This launch signals a broader industry transition toward AI-native infrastructure where software, rather than hardware cycles, dictates quality of service. Watch for whether rivals like Nokia can match these throughput gains using their alternative GPU-based acceleration strategies.
Additional Context
The launch of Ericsson's AI in RAN represents a critical junction in the architectural debate over how to modernize 5G networks. While Ericsson is championing an approach that utilizes its custom silicon and neural network accelerators embedded in its RAN Compute platforms, its primary western rival, Nokia, has taken a different path. Per Light Reading (March 2026), Nokia's strategy is heavily anchored in a partnership with NVIDIA, using GPU-accelerated computing for the most resource-intensive radio functions. This divergence forces operators to choose between Ericsson’s integrated, energy-efficient ASIC-based models or Nokia’s more flexible but potentially more expensive GPU-centric architecture. This competition is intensifying as network traffic patterns shift toward more data-heavy, asymmetric workloads. According to reporting from Fierce Network (February 2026), these AI-native features are specifically designed to address 'beyond human scale' complexity in network management, particularly for 5G Advanced and the emerging 6G standard. Earlier in 2026, Ericsson unveiled neural network accelerators in its Massive MIMO radios, which reportedly increased AI inference capability by tenfold (per Digit Watch, February 2026). This hardware foundation now supports the agentic AI capabilities found in the new software subscription. Furthermore, the broader ecosystem is consolidating around these intelligent networking standards via the AI-RAN Alliance. Founding members including Microsoft, Amazon, and Samsung are working to ensure that these software-driven optimizations can support external AI applications at the network edge. As noted by Telecompaper (June 2026), trials with T-Mobile in the U.S. have already demonstrated that AI-native scheduling can deliver stable performance for video calls and streaming even during peak congestion, validating the ROI for operators facing stagnant revenue despite rising bandwidth demands.
Read full article at techafricanews.com
Get this in your inbox → Subscribe
Enjoy our coverage?
Add StreamingMeme as a preferred source on Google to see more of our streaming news at the top of your Search results.
Add as preferred source