Content for TR 23.700-91 Word version: 17.0.0
7 Overall Evaluation p. 336
7.0 Summary of analytics provided by NWDAF p. 336
7.1 Key Issue #1: Logical decomposition of NWDAF and possible interactions between logical functions p. 338
7.2 Key Issue #2: Multiple NWDAF instances p. 339
7.4 Key Issue #4: Remaining aspects on how to ensure that slice SLA is guaranteed p. 347
7.7 Key Issue #7 Adding Application attributes and KPIs as the Input data in some services described in TS 23.288 [5] p. 348
7.8 Key Issue #8: UE data as an input for analytics generation p. 349
7.9 Key Issue #9: Dispersion Analytics output provided by NWDAF p. 350
7.10 Key Issue #10: NWDAF Assisted UP Optimization p. 351
7.11 Key Issue #11: Increasing efficiency of data collection p. 351
7.11.1 Evaluation for solutions based on parametrization and services changes p. 351
7.11.2 Evaluation for solutions based on Tracking and Discovery of Entities p. 352
7.11.3 Evaluation for solutions based on event exposure service enhancement p. 353
7.11.4 Signalling reduction via architectural changes p. 357
7.12 Key Issue #12: NWDAF-assisted RFSP policy p. 360
7.13 Key Issue #13: Triggering conditions for analytics p. 361
7.14 Key Issue #14: NWDAF-assisted application detection p. 364
7.15 Key Issue #15: User consent for UE data collection/analysis p. 364
7.16 Key Issue #16: UP optimization for edge computing p. 364
7.17 Key issue #17: Definition of accuracy levels p. 365
7.18 Key Issue #18: Enhancement for real-time communication with NWDAF p. 365
7.19 Key Issue #19: Trained data model sharing between multiple NWDAF instances p. 365
7.20 Key Issue #20: NWDAF assisting in detecting anomaly events for the user plane p. 366
7.21 Key Issue #21: NWDAF assisting in user plane performance p. 366
8 Conclusions p. 366
8.1 Key Issue #1: Logical decomposition of NWDAF and possible interactions between logical functions p. 366
8.2 Key Issue #2: Multiple NWDAF instances p. 367
8.2.1 Multiple NWDAF architecture and Analytics aggregation p. 367
8.2.2 Re-selection of NWDAF p. 368
8.2.3 Specific aspects of NWDAF interactions p. 369
8.4 Key Issue #4: Remaining aspects on how to ensure that slice SLA is guaranteed p. 369
8.7 Key Issue #7: Adding Application attributes and KPIs as the Input data in some services described in TS 23.288 [5] p. 370
8.8 Key Issue #8: UE data as an input for analytics generation p. 370
8.9 Key Issue #9: Dispersion Analytics output provided by NWDAF p. 371
8.10 Key Issue #10: NWDAF Assisted UP Optimization p. 372
8.11 Key Issue #11: Increasing efficiency of data collection p. 372
8.11.1 Signalling reduction via parametrization and services changes p. 372
8.11.2 Signalling reduction on Tracking and Discovery of Entities p. 372
8.11.3 Signalling reduction via Event Exposure service enhancement p. 372
8.11.4 Signalling reduction via architectural changes p. 373
8.12 Key Issue #12: NWDAF-assisted RFSP policy p. 374
8.13 Key Issue #13: Triggering conditions for analytics p. 375
8.15 Key Issue #15: User consent for UE data collection/analysis p. 375
8.16 Key Issue #16: UP optimization for edge computing p. 375
8.17 Key issue #17: Definition of accuracy levels p. 376
8.18 Key Issue #18: Enhancement for real-time communication with NWDAF p. 376
8.19 Key Issue #19: Trained data model sharing between multiple NWDAF instances p. 376
8.20 Key Issue #20: NWDAF assisting in detecting anomaly events for the user plane p. 376
8.21 Key Issue #21: NWDAF assisting in user plane performance p. 376
$ Change history p. 377
