3GPP 6G roadmap sets December 2028 functional freeze for Release 21

The formalization of Release 21 provides the streaming industry with a concrete timeline for the next leap in wireless delivery capacity. By locking in a December 2028 functional freeze, 3GPP is establishing the technical boundaries for ultra-low latency and high-bandwidth applications that will eventually succeed 5G-Advanced. For the streaming ecosystem, this roadmap dictates when engineers can begin designing for 6G-native features like immersive holographic video and massive-scale edge distribution. The industry should now monitor the specific radio deployment scenarios in TR 38.914 to understand how 6G will handle high-density video traffic in urban environments.

Ericsson has been among the most vocal equipment vendors framing its current AI-native RAN investments as stepping stones toward 6G. At MWC 2026, networks chief Per Narvinger outlined how AI models integrated into link adaptation algorithms can extract 10 percent more capacity from existing spectrum, a technique the company describes as foundational to the kind of autonomous optimization that 6G will require natively. Narvinger confirmed that by end of 2026 Ericsson will have 10 AI-ready radio products in market, each featuring custom Ericsson Silicon with embedded neural network accelerators designed for real-time on-site inference. That hardware cadence signals the vendor is building the silicon and software base that will eventually map onto Release 21 specifications once 3GPP finalizes the air interface.

On the business side, Ericsson's June 2026 Mobility Report provided the first quantified projections of how agentic AI workloads will reshape traffic profiles in ways that directly inform 6G requirements. The report found that a medium-quality AI agent running on AR headsets at 20 percent adoption could boost uplink traffic by 47 percent and downlink by 14 percent, figures that will feed into the spectrum and capacity assumptions underpinning TR 38.914's radio deployment scenarios. The report also distinguished between on-demand and always-on AI agents, noting that proactive autonomous agents will consume more network resources and require careful governance, a consideration likely to appear in 3GPP's service requirements for Release 21. Meanwhile, Ericsson published a technical blog in July 2025 detailing how agentic AI powered by AWS can reduce analysis and decision-making time by 80 percent, positioning the company's autonomous network ambitions at Level 5, the highest tier of self-optimization that 6G is expected to formalize.

Competitive positioning around 6G-ready capabilities is also emerging from the enterprise side. Cradlepoint, now part of Ericsson, integrated agentic AI into its NetCloud platform in 2025, becoming the first enterprise 5G vendor to do so, enabling autonomous task assignment across distributed router fleets without human intervention. That move demonstrates how agentic AI is already being operationalized at the network edge, providing real-world deployment data that 3GPP study items like TR 22.870 will reference when defining 6G service requirements. For streaming engineers tracking the 3GPP 6G roadmap, these vendor deployments offer early signals about the latency, uplink, and edge-compute parameters that Release 21 specifications will need to accommodate for immersive and AI-driven video experiences.