Content for TR 23.700-84 Word version: 19.0.0
1 Scope
2 References
3 Definitions of terms and abbreviations
3.1 Terms
3.2 Abbreviations
4 Architectural Assumptions and Requirements
...
...
1 Scope p. 10
This study will focus on the following objectives:
- AI/ML cross-domain coordination aspects on whether and how to consider 5GC enhancements to LCS to support AI/ML based Positioning considering conclusions of the RAN study in TR 38.843.
- Whether and what 5GC enhancements are needed to enable 5G system, including the AF, to assist in collaborative AI/ML operations involving NWDAF for Vertical Federated Learning (VFL).
- Whether and how to enhance 5GC to support NWDAF-assisted policy & QoS control.
- Whether and how to enhance 5GC to address network abnormal behaviour, i.e. signalling storm, with the assistance of NWDAF.
2 References p. 10
The following documents contain provisions which, through reference in this text, constitute provisions of the present document.
- References are either specific (identified by date of publication, edition number, version number, etc.) or non-specific.
- For a specific reference, subsequent revisions do not apply.
- For a non-specific reference, the latest version applies. In the case of a reference to a 3GPP document (including a GSM document), a non-specific reference implicitly refers to the latest version of that document in the same Release as the present document.
[1]
TR 21.905: "Vocabulary for 3GPP Specifications".
[2]
TS 23.501: "System Architecture for the 5G System; Stage 2".
[3]
TS 23.502: "Procedures for the 5G system, Stage 2".
[4]
TS 23.503: "Policy and Charging Control Framework for the 5G System".
[5]
TS 23.288: "Architecture enhancements for 5G System (5GS) to support network data analytics services".
[6]
TR 38.843: "Study on Artificial Intelligence (AI)/Machine Learning (ML) for NR air interface".
[7]
TS 23.273: "5G System (5GS) Location Services (LCS)".
[8]
TS 38.413: "NG Application Protocol (NGAP)".
[9]
TS 28.104: "Management and orchestration; Management Data Analytics (MDA)".
[10]
TS 22.071: "Technical Specification Group Systems Aspects; Location Services (LCS)".
[11]
TS 28.552: "Management and orchestration; 5G performance measurements".
[12]
TS 23.003: "Numbering, addressing and identification".
[13]
TS 29.510: "5G System; Network function repository services; Stage 3".
[14]
TS 28.554: "Management and orchestration; 5G end to end Key Performance Indicators (KPI)".
[15]
TS 38.300: "NR; NR and NG-RAN Overall description; Stage-2".
3 Definitions of terms and abbreviations p. 11
3.1 Terms p. 11
For the purposes of the present document, the terms given in TR 21.905 and the following apply. A term defined in the present document takes precedence over the definition of the same term, if any, in TR 21.905.
Horizontal Federated Learning (HFL):
a federated learning technique without exchanging/sharing local data set, wherein the local data set in different FL clients for local model training have the same feature space for different samples (e.g. UE IDs).
Vertical Federated Learning (VFL):
a federated learning technique without exchanging/sharing local data set, wherein the local data set in different VFL Participant for local model training have different feature spaces for the same samples (e.g. UE IDs).
Label:
A label is the training objective in supervised machine learning.
VFL Server:
An NWDAF or AF that integrates local training results, computes gradient information or loss information and send them to VFL client(s) for the local ML model update in VFL training process. It also coordinates the VFL training process by discovering and selecting VFL clients. In VFL inference process, The VFL server aggregates local inference results from VFL clients to generate the final VFL inference result and sends the final VFL inference result to the consumer. Only one VFL server may exist for each VFL process.
VFL Client:
An NWDAF or AF that holds the local dataset and performs local training and inference as asked by VFL Server. There can be multiple VFL Clients in VFL training and inference.
3.2 Abbreviations p. 11
For the purposes of the present document, the abbreviations given in TR 21.905 and the following apply. An abbreviation defined in the present document takes precedence over the definition of the same abbreviation, if any, in TR 21.905.
AP
Active Participant
HFL
Horizontal Federated Learning
PP
Passive Participant
VFL
Vertical Federated Learning
4 Architectural Assumptions and Requirements p. 11
The present study will not consider service-based interfaces with RAN and with UE.
The architecture for the present study shall comply with the existing NWDAF framework as specified in TS 23.288, and 5GS framework as specified in TS 23.501, TS 23.502 and TS 23.503.
The architecture for the present study shall comply with the existing Location Service Architecture as specified in TS 23.273.
Regarding AI/ML cross-domain coordination aspects, work will be based on the possible requirements defined by RAN WGs considering the conclusions in TR 38.843.
