Content for TR 22.876 Word version: 19.1.0
8.2 Consolidated potential KPIs
8.2.1 Split AI/ML operation between AI/ML endpoints
8.2.2 AI/ML model/data distribution and sharing by leveraging direct device connection
8.2.3 Distributed/Federated Learning by leveraging direct device connection
8.2 Consolidated potential KPIs p. 36
8.2.1 Split AI/ML operation between AI/ML endpoints p. 36
| Max allowed end-to-end latency (NOTE 1) | UL Payload size (Intermediate data size)
(NOTE 1) |
UL Experienced data rate
(NOTE 1) |
Service area dimension | Communication service availability
(NOTE 1) |
Reliability
(NOTE 1) |
Remarks |
|---|---|---|---|---|---|---|
| 2-100 ms | ≤1.5 Mbyte for each frame | ≤720 Mbps | Proximity-based work task offloading for Remote driving, AR displaying/gaming, remote-controlled robotics, video recognition and One-shot object recognition | |||
| 10 ms | ≤ 1.6 MByte (8 bits data format) | ≤1.28 Gbps | 900 m2 (30 m x 30 m) | 99.999 % | 99.99 % | Local AI/ML model split on factory robots |
| 10 ms | ≤ 6.4 Mbyte (32 bits data format) | ≤1.5 Gbps | Local AI/ML model split on factory robots | |||
|
NOTE 1:
The KPIs in the table apply to UL data transmission in case of indirect network connection.
|
||||||
8.2.2 AI/ML model/data distribution and sharing by leveraging direct device connection p. 37
| Max allowed end-to-end latency
(NOTE 1) |
Experienced data rate
(NOTE 1) |
Payload size
(NOTE 1) |
Communication service availability
(NOTE 1) |
Remark |
|---|---|---|---|---|
| 1s | ≤1.92 Gbit/s | ≤240 MByte | 99.9 % | AI Model Transfer Management through Direct Device Connection |
| 3s | ≤81.33 Mbyte/s | ≤244 MByte | - | transfer learning for trajectory prediction |
|
NOTE 1:
The KPIs in the table apply to data transmission using direct device connection.
NOTE 2:
The AI/ML model data distribution is for a specific application service.
|
||||
8.2.3 Distributed/Federated Learning by leveraging direct device connection p. 37
| Payload size
(NOTE 1) |
Maximum latency | Experienced data rate | Reliability | Remark |
|---|---|---|---|---|
| 132 MByte | 2-3 s | ≤528 Mbit/s | Direct device connection assisted Federated Learning (Uncompressed model) Asynchronous Federated Learning via direct device connection | |
| ≤50 MByte | 1 s | ≤220 Mbit/s | 99.99% | |
|
NOTE 1:
The KPIs in the table apply to both UL and DL data transmission in case of indirect network connection.
|
||||
