Tech-
invite
3GPP
space
IETF
space
21
22
23
24
25
26
27
28
29
31
32
33
34
35
36
37
38
4‑5x
Content for
TR 26.803
Word version: 17.0.0
0…
4…
4
Overview of Relevant Architectures
5
Use Cases for Edge Media Processing
6
Potential 5GMS Architecture Extensions
7
Conclusions and Recommendations
A
Deployment Scenarios
B
Selected Use Case Mapping
$
Change history
4
Overview of Relevant Architectures
p. 8
4.1
General
p. 8
4.2
SA6 Edge Architecture
p. 8
4.2.1
Edge computing architecture
p. 8
4.2.2
High-level call flows for EAS relocation scenarios
p. 13
4.3
SA2 Edge Support
p. 15
4.3.1
Overview
p. 15
4.3.2
Process of Discovering EAS
p. 17
4.3.2.1
General
p. 17
4.3.2.2
DNS-based solutions for Multiple PDU Sessions
p. 17
4.3.2.3
DNS-based solutions for Distributed Anchors
p. 18
4.3.2.4
DNS-based solutions for Session Breakout
p. 18
4.3.3.5
Summary of DNS based EAS Discovery
p. 18
4.3.3.6
Discussion
p. 18
4.3.4
Summary of EAS relocation solutions
p. 19
4.4
SA5 Edge Management
p. 19
4.5
5GMS Architecture
p. 20
5
Use Cases for Edge Media Processing
p. 21
5.1
General
p. 21
5.2
Downlink Streaming Use Cases
p. 21
5.2.1
Caching Downlink Streaming Content
p. 21
5.2.2
Split Rendering
p. 22
5.2.3
User-generated live streaming
p. 25
5.2.4
Augmented Video Streaming
p. 29
5.2.5
Generalized Split and Cloud Rendering and Processing
p. 31
5.2.6
Photo-realistic AR Rendering in Network
p. 32
5.2.7
Media Services in the Edge
p. 34
5.2.8
Partial delivery of 3D content (point cloud, mesh) for AR/MR device
p. 35
5.3
Uplink Streaming Use Cases
p. 36
5.3.1
Multi-Camera Uplink Stream Processing
p. 36
5.3.2
Cloud/Split Rendering of Immersive Live Events
p. 39
5.3.3
Pandemic Stadium
p. 40
5.4
Analysis of Use Cases
p. 42
5.4.1
Summary
p. 42
5.4.2
Use Cases requiring only downlink media streaming
p. 43
5.4.2.1
Caching downlink streaming content
p. 43
5.4.2.2
Split rendering
p. 43
5.4.2.3
Generalized split and cloud rendering and processing
p. 43
5.4.2.4
Cloud/split rendering of immersive live events
p. 44
5.4.2.5
Pandemic stadium
p. 44
5.4.2.6
Media services in the edge
p. 44
5.4.3
Use Cases requiring only uplink media streaming
p. 44
5.4.4
Use Cases requiring both uplink and downlink media streaming
p. 44
5.4.4.1
User-generated live streaming
p. 44
5.4.4.2
Augmented video streaming
p. 44
5.4.4.3
Photo-realistic AR rendering in network
p. 44
5.4.4.4
Partial delivery of 3D content (point cloud, mesh) for AR/MR device
p. 44
5.4.4.5
Multi-camera uplink stream processing
p. 45
6
Potential 5GMS Architecture Extensions
p. 45
6.1
General
p. 45
6.2
EMSA Architecture
p. 45
6.3
Generic Call Flows for Media Session Establishment
p. 46
6.3.1
General
p. 46
6.3.2
Client-driven edge discovery
p. 47
6.3.3
AP-driven management of 5GMS edge processing
p. 50
6.3.4
Application Context Relocation for EAS instances
p. 51
6.3.4.1
General
p. 51
6.3.4.2
Scenario 1: EAS relocation decided by EEC
p. 53
6.4
Identified Gaps in Architecture and Procedures
p. 54
6.4.1
General
p. 54
6.4.2
Justification of Identified Gaps
p. 55
6.4.2.1
Gaps in client-driven edge discovery
p. 55
6.4.2.2
Gaps in AP-driven management of 5GMS edge processing
p. 55
6.4.2.3
Gaps in Application Context Relocation
p. 55
7
Conclusions and Recommendations
p. 55
A
Deployment Scenarios
p. 57
A.1
Deployment Scenario Option A
p. 57
A.2
Deployment Scenario Option B
p. 58
B
Selected Use Case Mapping
p. 59
B.1
Split Rendering
p. 59
B.2
Edge Caching
p. 61
B.3
User-generated live streaming
p. 62
B.3.1
Service provisioning
p. 62
B.3.1.1
Overview
p. 62
B.3.1.2
Example media processing features
p. 62
B.3.1.3
Example resource template aspects
p. 63
B.3.2
High-level call flow
p. 64
B.3.3
Service relocation considerations
p. 65
$
Change history
p. 66