VFL Training Trigger Conditions and Way Forward Discussion Summary

Triggering VFL training Way Forwards 1

For trigger condition for VFL training or VFL inference, we had a lots of discussion in Jan emeeting and S2-2501134 from Ericsson, etc had been technically endorsed; Several companies brought discussion papers about trigger condition for VFL training or VFL inference S2-2501477(Nokia), 7 cases and corresponding solutions are discussed. S2-2501483(Ericsson), discussion to agree the endorsed CR involving 3 cases; S2-2501547(Lenovo), 3 cases similar as the ones in Ericsson s CR for VFL training are proposed and suggested to capture in TS; S2-2502035(HW), 3 cases similar as the ones in Ericsson s CR are mentioned, but proposed to document them in the clause of inference. S2-2501984(ZTE), 5 cases are mentioned, 1-3 cases are basic 4-5 are also OK to capture. Proposal 1: Step1: discuss which cases can be agreed first from technical point of view. Step2: Where and how to document them can be decided based on the cases agreed. 2

Case 0: VFL training triggered by OAM configuration or internal logic VFL server NWDAF (MTLF+AnLF) VFL Clients 3. Internal logic/OAM configuration 4.VFL training Way Forward: Since no one disagreed this case, we may keep it in TS, more explicit clarification on wording may be needed. Suggested wording: XX (or discussed during the meeting) 3

Case 1: VFL training triggered by analytics request from NF or AnLF B) NF consumer discovers AnLF, AnLF discovers VFL server (NWDAF or AF) A) NF consumer discovers a VFL server (NWDAF) directly VFL server (AF or NWDAF ) Proxy NWDAF containing AnLF Analytics Consumer (e.g. PCF) VFL server NWDAF (MTLF+AnLF) Consumer NF VFL Clients NEF 0. Nnwdaf_AnalyticsInfo_Request / Nnwdaf_AnalyticsSubscription_Subscribe 1. discovers NWDAF(AnLF) at NRF to provide Analytics VFL clients 2. Nnwdaf_Analytics_Subscribe 1: AnLF is enhanced to discover (e.g. or be aware of) VFL server to provide analytics from NRF 3. If no model available, perform VFL training 2. Nnwdaf/Af/nef_AnalyticsInfo_Request / Nnwdaf_AnalyticsSubscription_Subscribe 4.VFL training 5.VFL inference 6. Nnwdaf_Analytics_Notify 3. Trigger VFL training 4. VFL inference Back ground info: 1-A is corresponding to case 0d in endorsed paper, and implicitly supported in inference part (6.2H.2.4); 1-B is implicitly supported in inference part (6.2H.2.4); Technically, no one disagree this case 1 is one of trigger for VFL training; Some company concerned where to document this case, i.e. in VFL training clause or in VFL inference clause, or both? Because this case indeed involve both VFL training and VFL inference processes and combined them together. Ericsson requests offline to clarify that: a specific analytics uses only VFL and the AnLF knows that Nokia reply: This is visible in the NF profile retrieved by the AnLF (VFL server capability), so AnLF does not need to know in advance 5a. Nnwdaf_AnalyticsInfo_Response Way forwards: keep this case in TS, but discuss where and how to document with some essential clarification if needed 4

Case 2: VFL training at NWDAF VFL server triggered by model retrieval request from AnLF A) To retrieve ML model, AnLF discovers a VFL server (NWDAF) directly B) To retrieve ML model, AnLF discovers MTLF, then MTLF discovers a VFL server (AF) NWDAF that acts as VFL inference server Traditional NWDAF (AnLF) NWDAF (MTLF) (is an VFL server) NWDAF that acts as VFL training server VFL clients 0.Nnwdaf_analyticsinfo_ subscribe/request 1. Determine that model is required, e.g. based on configuration, analytics monitoring or inference request 1. Decides to retrieve an ML model from MTLF, 2. Nnwdaf_MLModelProvision_Subscribe 2.Nnwdaf_MLmodelProvision_Subscribe No model 3. Train VFL model 4. Nnwdaf_MLModelProvision_Notify(VFL server model retrieval info, VFL correlation ID, vendor container VFL context(weights, clients, features etc) 3.Starts VFL training VFL training comple Back ground info: 2-A is corresponding to case 0c in endorsed paper. Wherein, step 1 may be decided based on AnLF s internal requirement of ML model or based on analytics request from analytics consumer (i.e. dash line in step 0). Case 2_B is normal model retrieval procedure since Rel-17 and also as used for HFL, but with extensions to enable split VFL server People have different view on this case 2, some companies can not accept this case 2A. Some one comment even this case is accepted, they should be documented into clause for VFL inference. way forwards: Step 1: discuss whether we need case A and/or B or not. Step 2: if yes, then decide where to document and how. (please see next slide about potential documentation proposal ) 5

Case 3: VFL training at AF VFL server supervised by AnLF A) To retrieve ML model, AnLF discovers a VFL server (NWDAF) directly B) To retrieve ML model, AnLF discovers MTLF, then MTLF discovers a VFL server (AF) Traditional NWDAF (AnLF) Proxy NWDAF (MTLF) AF VFL clients (VFL server) AF that acts as VFL training server Supervising AnLF 0.Nnwdaf_analyticsinfo_ subscribe/request 1. Determine that modelmodel training is required 1. Decides to retrieve an ML model from MTLF, 2. Naf_Model_Training request 2.Nnwdaf_Mlmode lProvision_Subscribe 3.VFL training 3. Train VFL model 4.Starts VFL training 5a.VFL training complete 4. Model trainiong complete 5b.VFL training complete Back ground info: 2-A is simplified proposal assuming that AnLF that forwards inference request to AF can also supervise training 2-B is corresponding to case 0a+0b in endorsed paper. People have different view on this case 2B, some companies can not accept this case 2B Some one comment even case 2B is accepted, it should be documented into clause for VFL inference. way forwards: Step 1: discuss whether we need this case or not and select solution A or B Step 2: if yes, then decide where to document and how. (please see next slide about potential documentation proposal ) 6

Ericssons proposal on documentation and rewording: For 2-A (corresponding to case 0c in the endorsed paper): in training 0c was added for the case where the NWDAF is the VFL Server, (not the AF), then AnLF contacts MTLF that happens to be the VFL Server and can start training with the VFL Clients. For 2-B (corresponding to case 0a, 0b in the endorsed paper : in training & in inference 1. In training procedure, we add the possibility in step 0a and b to add an optionalflag indicating inference shall be started when training is done and result shall be sent to consumer . This is to not force AnLF to request Inference if this is part of Inference procedure. 2. In Inference procedure a reference to step 0a and b is done, stating that the flag shall be set. Please Ericsson can clarify further to help people to about this proposal on this page. The call flow revised by Ericsson is shown in the next page. 7

NWDAF (AnLF) NWDAF (MTLF) NWDAF (VFL client) Trusted AF (VFL client) Untrusted AF (VFL client) NF NEF VFL server ... ... 0a. Nnwdaf_MLModelProvision_Subscribe Response to subscription [MTLF decides either 0b) respond no model will be available or oc) ask AnLF to request Training directly to FL server] NWDAF (VFL client) Untrusted AF (VFL client) 0b. Naf_Training_Subscribe 0c. Naf_Training_Subscribe 0d. Nnwdaf_MLModelProvision_Subscribe [ask for VFL no model will be shared] 0e.Nnwdaf_AnalyticsSubscription_Subscribe Response to subscription 1. VFL Client Selection and VFL preparation phase Repeat until training completes 2. Start VFL training / Start another VFL training round 2a. Nnwdaf_VflTraining_Subscribe 2b. Naf_VflTraining_Subscribe 2c. Nnef_VflTraining_AFClient_Subscribe 2d. Naf_Vfl_Training_Subscribe 3. Collect input data 4. Local model Update 5. Local model information reporting 5a. Nnwdaf_MLModel_Training_Notify or Nnwdaf_MLModelTrainingInfo_Request response. 5b. Naf_Vfll_Training_Notify 5c. Naf_Vfl_Training_Notify 5d. Nnef_Vfl_Training_AFClient_Notify 6. Combine intermediate results and calculate training results 7a. Evaluate results 7b. Naf_VFLTraining_Notification 7c. Nnwdaf_MLModelProvision_Subscribe 7d. Nnwdaf_MLModelProvision_Subscribe/unsubscribe 7e. Naf_VFLTraining_Subscribe/unsubscribe 7f. Nnwdaf_MLModelProvision_Notification 7g. Nnwdaf_MLModelProvision_Subscribe/unsubscribe 7h. Update local model and continue or terminate the VFL training process 8. VFL training Termination 8a. Nnwdaf_VFLTraining_Unsubscribe 8b. Naf_VflTraining_Unsubscribe 8c. Nnef_VflTraining_AFClient_Unsubscribe 8d. Naf_Vfl_Training_Unubscribe 9. store model 9. store model 8