Skip to main content
Testbeds

AI/ML Evaluation Framework

Evaluation framework for assessing AI/ML models applied to media processing in 5G network contexts.

The tools provide standardised benchmarks and datasets for evaluating machine learning models for media quality enhancement, adaptive bitrate (ABR) optimisation, and traffic classification, aligned with the AI/ML work items of 3GPP working group SA4 (the group responsible for codec and media work). It is intended for researchers and engineers contributing to the standardisation of AI-native media delivery and informing 5G-MAG's 6G research direction. For acronyms used here, see the Glossary.

The specifications below are listed as background context, not as a prerequisite reading list.

Specifications referenced: 3GPP TR 26.847 (Evaluation of AI and ML in 5G media services, Release 19, completed; supersedes the earlier, incomplete draft tracked as TR 26.927), TR 22.874 (AI/ML in 5G scenarios), aligned with 3GPP SA4 AI/ML study items in Release 18/19.

note

The reference tooling is currently private and under testing. Request early access.

Contributors

Project Overview

The slide deck below introduces the AI/ML Evaluation Framework: what it evaluates (media quality enhancement, adaptive bitrate optimisation, and AI traffic classification), the benchmarks and datasets it provides, and how it relates to the shared 6G Testbed used for the hands-on traffic characterisation work.

Download the slide deck of this Project

Go deeper

AreaDetails
ScopeFocus areas, repositories, and relationship to standards: Scope
StandardsAI/ML in 5G media standardisation: AI/ML on the Standards portal
Related projectAI traffic classification testbed: 6G Testbed and AI Traffic
In Action

Developer Exchange

No AI/ML Evaluation Framework-specific videos are curated here yet. In the meantime, here's what's recent from the wider Developer Exchange:

In Action

End-to-End setup with 5GC, NG-RAN and UE for MBS Broadcast Services - 5G-MAG Reference Tools

In Action

Overview of the latest updates to the 5G Media Streaming Project - 5G-MAG Reference Tools

In Action

Demonstrating MBS User Service Announcement mechanisms - 5G-MAG Reference Tools

In Action

Introducing Network Emulator & AI Traffic Characterization Framework - 5G-MAG Reference Tools

In Action

Introducing the new 5GMS Application Provider Portal - 5G-MAG Reference Tools

Browse all 5G-MAG videos →

Community Stats

Last Synced: 2026-07-28 01:13 UTC

RepositoryStarsForksViews (Total)Clones (Total)Activity
rt-ai-ml-evaluation-framework121070Stable