6G AI Traffic Characterization is a testbed of the AI in Media Services project.
This study aims to identify media-related opportunities and gaps in the context of 6G. It builds on service requirements defined by SA1 and architectural enhancements defined by SA2. The conclusions of this study will form the basis for further detailed studies as well as normative work.
Annexes C and D of the study document the evaluation of AI traffic characteristics and the testbed for it, which is 6G AI Traffic Characterization.
This project is based on technology developed by:
The Problem It Solves
The traffic characteristics of media services are relevant to the design of 6G radio and service architectures, and for AI media services the study obtains them itself. Its five initial findings are that AI traffic is uplink heavy in some scenarios, that it is bursty, that round-trip delay determines responsiveness, that traffic varies within one application, and that some traffic is structured in tokens. The findings are to be updated as more measurement results become available, which is what the testbed produces.
The project at a glance
What it is, how it works, and where the Reference Tools stand, in one deck you can share.
Download the deck (6G AI Traffic Characterization)From idea to running code
Read it, check it against the standard, then run it.
Specifications
The specifications this technology is built on.
Latest updates and roadmap
Status on 23 October 2026
for further study
- MetricsWhether and how TTFT, TTLT and tool-call latency can be used for AI traffic evaluation is for further study in TR 26.870.
- Agentic uplinkWhether the retransmitted context that explains the uplink-heavy agentic traffic occurs in agentic applications is for further study.
- AI gatewayThe test scenario of the AI gateway, and the source code of its call flow, are still to be provided.
in development
- Agentic scenariosPull request 10 adds the agentic measurement scenarios agreed for the 6G media study and separates network measurement into the standalone netemu package.
- Video analysis scenariosScenario configurations for video-traffic analysis.
- Chat with token IDsA chat scenario configuration in which the client sends token identifiers.
Open Source, Built Together
Repositories of This Project
The real, running software for this technology: what each repository is, what it implements, its latest release and where to find it.
Early Access: available to members first
6G-TestbedJoin the Effort
Help shape what connected media looks like next and scale your projects with the industry.