ITU-T Work Programme

[2025-2028] : [SG20] : [Q4/20]

[Declared patent(s)] - [Associated work]

Work item:

Y.AIoT-dfs-arc

Subject/title:

Reference architecture of data fusion service in artificial intelligence of things

Status:

Under study [Issued from previous study period]

Approval process:

AAP

Type of work item:

Recommendation

Version:

New

Equivalent number:

Timing:

2027-Q1 (Medium priority)

Liaison:

ITU-T SG2, SG11, SG13, SG17

Supporting members:

China Unicom; ZTE Corporation; ETRI; China Information Communication Technologies Group; Beijing University of Posts and Telecommunications; China Telecommunications Corporation

Summary:

Data fusion service (DFS) is a converged data service in artificial intelligence of things (AIoT). It merges, combines and integrates data and metadata from diverse resources — including IoT devices, edge systems, databases, cloud platforms — into a coherent, interoperable form. By leveraging the knowledge database (such as relational, graph), as well as artificial intelligence (AI) and machine learning (ML) capabilities, DFS provides advanced data processing and management for consumers, such as those performing machine learning model training and service orchestration. The DFS dynamically collects metadata to generate intelligent data catalogs in real-time. It provides a smart data provision capability that ensures data consumers can access assets while meeting specific quality of experience (QoE) and quality of service (QoS) requirements. Furthermore, DFS enables organizations to unify disparate data systems, strengthen data governance, and provide seamless, intelligent data accessibility across multiple platforms to enhance cross-domain collaboration. This draft Recommendation specifies the DFS in AIoT, detailing its technical characteristics, functional requirements, reference architecture, and core capabilities.

Comment:

Reference(s):

[SG20-TD1515/GEN (2026-05)

]