hash stringlengths 32 32 | doc_id stringlengths 5 12 | section stringlengths 5 1.47k | content stringlengths 0 6.67M |
|---|---|---|---|
fbecc7f0dcf9784c6066646052ab0c0e | 23.501 | 5.17.2.3.2 Mobility for UEs in single-registration mode | When the UE supports single-registration mode and network supports interworking procedure without N26 interface:
- For mobility from 5GC to EPC, the UE with at least one PDU Session established in 5GC may either:
- if supported and if it has received the network indication that interworking without N26 is supported, perform Attach in EPC with a native EPS GUTI, if available, otherwise with IMSI with Request type "Handover" in PDN CONNECTIVITY Request message (clause 5.3.2.1 of TS 23.401 [26]) and indicating that the UE is moving from 5GC and subsequently moves all its other PDU Session using the UE requested PDN connectivity establishment procedure with Request Type "handover" flag (clause 5.10.2 of TS 23.401 [26]), or.
- perform TAU with 4G-GUTI mapped from 5G-GUTI sent as old Native GUTI (clause 5.3.3 of TS 23.401 [26]) indicating that it is moving from 5GC, in which case the MME instructs the UE to re-attach. IP address preservation is not provided in this case.
- for the first TAU after 5GC initial Registration, the UE and MME for the handling of UE Radio Capabilities follow the procedures as defined in clause 5.11.2 TS 23.401 [26] for first TAU after GERAN/UTRAN Attach.
NOTE 1: The first PDN connection may be established during the E-UTRAN Initial Attach procedure (see TS 23.401 [26]).
NOTE 2: At inter-PLMN mobility to a PLMN that is not an equivalent PLMN the UE always uses the TAU procedure.
- For mobility from 5GC to EPC, the UE with no PDU Session established in 5GC
- performs Attach in EPC (clause 5.3.2.1 of TS 23.401 [26]) indicating that the UE is moving from 5GC.
- For mobility from EPC to 5GC, the UE performs Mobility Registration Update in 5GC with 5G-GUTI mapped from EPS GUTI and a native 5G-GUTI, if available, as Additional GUTI and indicating that the UE is moving from EPC. In this case, the AMF determines that old node is an MME, but proceeds as if the Registration is of type "initial registration". The UE may either:
- if supported and if it has received the network indication "interworking without N26 supported", move all its PDN connections from EPC using the UE initiated PDU Session Establishment procedure with "Existing PDU Sessions" flag (clause 4.3.2.2.1 of TS 23.502 [3]), or
- re-establish PDU Sessions corresponding to the PDN connections that it had in EPS. IP address preservation is not provided in this case.
NOTE 3: The additional native 5G-GUTI enables the AMF to find the UE's 5G security context (if available).
NOTE 4: When single-registration mode UE uses interworking procedures without N26, the registration states during the transition period (e.g. while UE is transferring all PDU Sessions / PDN Connections on the target side) are defined in Stage 3 specifications.
- If the network determines that the UE is changing RAT type, if the UE requests to relocate the PDU session from EPC to 5GC or 5GC to EPC, the SMF/MME uses the "PDU session continuity at inter RAT mobility" or "PDN continuity at inter-RAT mobility" information, respectively, in the subscription to determine whether to maintain the PDU session/PDN connection (if being handed over) or reject the PDU session request, with the relevant cause.
- If the UE requested to move the PDU session and the "PDN continuity at inter RAT mobility" information indicated "disconnect the PDN connection with a reactivation request" the network should provide a suitable cause code to the UE so that it can request a new PDU session. |
fbecc7f0dcf9784c6066646052ab0c0e | 23.501 | 5.17.2.3.3 Mobility for UEs in dual-registration mode | To support mobility in dual-registration mode, the support of N26 interface between AMF in 5GC and MME in EPC is not required. A UE that supports dual registration mode may operate in this mode when it receives an indication from the network that interworking without N26 is supported.
For UE operating in dual-registration mode the following principles apply for PDU Session transfer from 5GC to EPC:
- UE operating in Dual Registration mode may register in EPC ahead of any PDU Session transfer using the Attach procedure indicating that the UE is moving from 5GC without establishing a PDN Connection in EPC if the EPC supports EPS Attach without PDN Connectivity as defined in TS 23.401 [26]. Support for EPS Attach without PDN Connectivity is mandatory for UE supporting dual-registration procedures.
NOTE 1: Before attempting early registration in EPC the UE needs to check whether EPC supports EPS Attach without PDN Connectivity by reading the related SIB in the target cell.
- UE performs PDU Session transfer from 5GC to EPC using the UE initiated PDN connection establishment procedure with "handover" indication in the PDN Connection Request message (clause 5.10.2 of TS 23.401 [26]).
- If the UE has not registered with EPC ahead of the PDU Session transfer, the UE can perform Attach in EPC with "handover" indication in the PDN Connection Request message (clause 5.3.2.1 of TS 23.401 [26]).
- UE may selectively transfer certain PDU Sessions to EPC, while keeping other PDU Sessions in 5GC.
- UE may maintain the registration up to date in both 5GC and EPC by re-registering periodically in both systems. If the registration in either 5GC or EPC times out (e.g. upon mobile reachable timer expiry), the corresponding network starts an implicit detach timer.
NOTE 2: Whether UE transfers some or all PDU Sessions on the EPC side and whether it maintains the registration up to date in both EPC and 5GC can depend on UE capabilities that are implementation dependent. The information for determining which PDU Sessions are transferred on EPC side and the triggers can be pre-configured in the UE and are not specified in this Release of the specification. The UE does not know before-hand, i.e. before trying to move a given PDU session to EPC, whether that PDU session can be transferred to EPC.
NOTE 3: The Start of Unavailability Period and/or Unavailability Period Duration that the UE determines for NR satellite access with discontinuous network coverage in 5GS (see clause 5.4.1.4) and determines for satellite access with discontinuous coverage in EPS (see clause 4.13.8.2 of TS 23.401 [26]) can be different between 5GS and EPS.
For UE operating in dual-registration mode the following principles apply for PDN connection transfer from EPC to 5GC:
- UE operating in Dual Registration mode may register in 5GC ahead of any PDN connection transfer using the Registration procedure indicating that the UE is moving from EPC (clause 4.2.2.2.2 of TS 23.502 [3]).
- UE performs PDN connection transfer from EPC to 5GC using the UE initiated PDU Session Establishment procedure with "Existing PDU Session" indication (clause 4.3.2.2.1 of TS 23.502 [3]).
- UE may selectively transfer certain PDN connections to 5GC, while keeping other PDN Connections in EPC.
- UE may maintain the registration up to date in both EPC and 5GC by re-registering periodically in both systems. If the registration in either EPC or 5GC times out (e.g. upon mobile reachable timer expiry), the corresponding network starts an implicit detach timer.
NOTE 4: Whether UE transfers some or all PDN connections on the 5GC side and whether it maintains the registration up to date in both 5GC and EPC can depend on UE capabilities that are implementation dependent. The information for determining which PDN connections are transferred on 5GC side and the triggers can be pre-configured in the UE and are not specified in this Release of the specification. The UE does not know before-hand, i.e. before trying to move a given PDN connection to 5GC, whether that PDN connection can be transferred to 5GC.
NOTE 5: If EPC does not support EPS Attach without PDN Connectivity the MME detaches the UE when the last PDN connection is released by the PGW as described in clause 5.4.4.1 of TS 23.401 [26] (in relation to transfer of the last PDN connection to non-3GPP access).
When sending a control plane request for MT services (e.g. MT SMS) the network routes it via either the EPC or the 5GC. In absence of UE response, the network should attempt routing the control plane request via the other system.
NOTE 6: The choice of the system through which the network attempts to deliver the control plane request first is left to network configuration. |
fbecc7f0dcf9784c6066646052ab0c0e | 23.501 | 5.17.2.3.4 Redirection for UEs in connected state | When the UE supports single-registration mode or dual-registration mode without N26 interface:
- If the UE is in CM-CONNECTED state in 5GC, the NG-RAN may perform RRC Connection Release with Redirection to E-UTRAN based on certain criteria (e.g. based on local configuration in NG-RAN, or triggered by the AMF upon receiving Handover Request message from NG-RAN).
- If the UE is in ECM-CONNECTED state in EPC, the E-UTRAN may perform RRC Connection release with redirection to NG-RAN based on certain criteria (e.g. based on local configuration in E-UTRAN, or triggered by the MME upon receiving handover request from E-UTRAN). |
fbecc7f0dcf9784c6066646052ab0c0e | 23.501 | 5.17.2.4 Mobility between 5GS and GERAN/UTRAN | IP address preservation upon direct mobility between 5GS and GERAN/UTRAN is not supported.
Upon mobility from 5GS to GERAN/UTRAN (e.g. upon leaving NG-RAN coverage) the UE shall perform the A/Gb mode GPRS Attach procedure or Iu mode GPRS Attach procedure (see TS 23.060 [56]).
As a UE option, to support IP address preservation at mobility from EPC to 5GS for PDN connections without 5GS related parameters, a 5GS capable UE may:
- Following mobility from GERAN/UTRAN to EPS, release those PDN connection(s) and re-establish them as specified in clause 4.11.1.5.4.1 of TS 23.502 [3] so that they support interworking to 5GS.
NOTE 1: It is recommended that a UE using this option does not do this behaviour after every change to EPS in PLMNs that do not support 5GS, nor for APNs that do not support mobility to 5GS; and, that such a UE supports storage of the 5GS related parameters while in GERAN/UTRAN. Whether and how the UE is aware of which PLMNs support 5GS and which APNs do not support mobility to 5GS is out of scope of this specification.
To support mobility from EPC to 5GS to EPC to GERAN/UTRAN for PDN connections established in EPC:
NOTE 2: For the use of N7 or N40 interfaces while the UE is in GERAN/UTRAN access, the SMF+PGW-C selected by the MME (using the existing selection procedures described in clause 4.11.0a of TS 23.502 [3] and clause 4.3.8 of TS 23.401 [26]) needs to support functionality (e.g. signalling of GERAN/UTRAN cell identification over N7) specified in Annex L.
- in signalling sent on the N26 interface, the MME should send the TI and BSS Container in the EPS Bearer Context (see Table 7.3.1-3 of TS 29.274 [101]), if there is any, of the EPS bearer to the SMF (V-SMF / I-SMF) via the AMF in the Bearer Context within the PDN Connection IE in the Forward Relocation Request and Context Response messages (TS 29.274 [101]); the SMF (V-SMF / I-SMF) should store the TI and BSS Container and the SMF (V-SMF / I-SMF) should provide the TI and BSS Container to the AMF (as part of a procedure to deliver SM context to AMF) so that the AMF sends the TI and BSS Container of the related EPS bearer in the Bearer Context within the EPS PDN Connection information in any subsequent Forward Relocation Request and Context Response message sent to an MME.
NOTE 3: At mobility from EPC, the SMF+PGW-C / V-SMF / I-SMF receives the TI and BSS Container as part of the UE EPS PDN Connection information from the AMF and stores the TI. At mobility to EPC, the SMF+PGW-C / V-SMF / I-SMF provides the AMF with the TI and BSS Container as part of the UE EPS PDN Connection information. The SMF+PGW-C / V-SMF / I-SMF is not meant to understand the TI/BSS Container nor to use it for any other purpose than providing it back to AMF.
GERAN/UTRAN Mobility Management Bearer Synchronisation procedures will release any dedicated QoS Flows established in 5GS and then the SMF+PGW-C deletes the PCC rule(s) associated with those QoS Flows and informs the PCF about the removed PCC rule(s) as described in clause 4.11 of TS 23.502 [3].
When the UE accesses the network via GERAN/UTRAN over Gn/Gp interface, Secondary PDP Context Activation Procedure is not supported and the SMF+PGW-C rejects the Secondary PDP Context Activation request if UE requested it.
IP address preservation at mobility from EPC to GERAN/UTRAN for PDU sessions established in 5GS is not supported.
With regard to interworking between 5GS and the Circuit Switched domain when the GERAN or UTRAN network is operating in NMO II (i.e. no Gs interface between MSC and SGSN):
- upon mobility from 5GS to GERAN/UTRAN, the UE shall either:
- act as if it is returning after a loss of GERAN/UTRAN coverage (and e.g. only perform a periodic LAU if the periodic LAU timer has expired), or,
- perform a Location Update to the MSC. If the UE is registered for IMS voice and is configured, using Device Management or initial provisioning, to perform additional mobility management procedures when it has moved from a RAT that supports IMS voice over PS sessions to one that does not (see TS 23.060 [56]), it shall follow this option.
Upon mobility from GERAN/UTRAN to 5GS (e.g. upon selecting an NG-RAN cell) the UE shall perform the Registration procedure of "initial registration" type as described in TS 23.502 [3]. The UE shall indicate a 5G-GUTI as UE identity in the Registration procedure if it has a stored valid native 5G-GUTI (e.g. from an earlier registration in the 5G System). Otherwise the UE shall indicate a SUCI.
If a UE in MICO mode moves to GERAN/UTRAN and any of the triggers defined in clause 5.4.1.3 occur, then the UE shall locally disable MICO mode and perform the A/Gb mode GPRS Attach procedure or Iu mode GPRS Attach procedure (see TS 23.060 [56]). The UE can renegotiate MICO when it returns to 5GS during (re-)registration procedure.
In Single Registration mode, expiry of the periodic RAU timer, or, the periodic LAU timer shall not cause the UE to change RAT.
The 5G SRVCC from NG-RAN to UTRAN is specified in the TS 23.216 [88]. After the 5G SRVCC to UTRAN, all the PDU sessions of the UE are released. |
fbecc7f0dcf9784c6066646052ab0c0e | 23.501 | 5.17.2.5 Secondary DN authentication and authorization in EPS Interworking case | Secondary authentication/authorization by a DN-AAA server during the establishment of a PDN connection over 3GPP access to EPC, is supported based on following principles:
- It is optional for the UE to support EAP-based secondary authentication and authorization by DN-AAA over EPC,
- A SMF+PGW-C shall be used to serve DNN(s) requiring secondary authentication/authorization by a DN-AAA server,
- For secondary authentication/authorization by a DN-AAA server, the SMF+PGW-C runs the same procedures with PCF, UDM and DN-AAA and uses the same corresponding interfaces regardless of whether the UE is served by EPS or 5GS,
- The interface towards the UE is different (usage of NAS for EPS instead of NAS for 5GS) between the EPS and 5GS cases.
This is further specified in Annex H of TS 23.502 [3].
In this Release, EAP based Secondary authentication by a DN-AAA server during the establishment of a PDN connection over non-3GPP access to EPC is not supported. |
fbecc7f0dcf9784c6066646052ab0c0e | 23.501 | 5.17.3 Interworking with EPC in presence of Non-3GPP PDU Sessions | When a UE is simultaneously connected to the 5GC over a 3GPP access and a non-3GPP access, it may have PDU Sessions associated with 3GPP access and PDU Sessions associated with non-3GPP access. When inter-system handover from 5GS to EPS is performed for PDU Sessions associated with 3GPP access, the PDU Sessions associated with non-3GPP access are kept anchored by the network in 5GC and the UE may either:
- keep PDU Sessions associated with non-3GPP access in 5GS (5GC+N3IWF or TNGF) (i.e. the UE is then registered both in EPS and, for non-3GPP access, in 5GS); or
- locally or explicitly release PDU Sessions associated with non-3GPP access; or
- once in EPS, transfer PDU Sessions associated with non-3GPP access to E-UTRAN by triggering PDN connection establishment with Request Type "Handover", as specified in TS 23.401 [26]. |
fbecc7f0dcf9784c6066646052ab0c0e | 23.501 | 5.17.4 Network sharing support and interworking between EPS and 5GS | The detailed description for supporting network sharing and interworking between EPS and 5GS is described in clauses 4.11.1.2.1, 4.11.1.2.2, 4.11.1.3.2 and 4.11.1.3.3 of TS 23.502 [3]. |
fbecc7f0dcf9784c6066646052ab0c0e | 23.501 | 5.17.5 Service Exposure in Interworking Scenarios | |
fbecc7f0dcf9784c6066646052ab0c0e | 23.501 | 5.17.5.1 General | Clause 4.3.5 shows the Service Exposure Network Architecture in scenarios where for EPC-5GC Interworking is required.
In scenarios where interworking between 5GS and EPC is possible, the network configuration is expected to associate UEs with SCEF+NEF node(s) for Service Capability Exposure. The SCEF+NEF hides the underlying 3GPP network topology from the AF (e.g. SCS/AS) and hides whether the UE is served by 5GC or EPC.
If the service exposure function that is associated with a given service for a UE is configured in the UE's subscription information, then an SCEF+NEF identity shall be used to identify the exposure function. For example, if a UE is capable of switching between EPC and 5GC, then the SCEF ID that is associated with any of the UE's APN configurations should point to an SCEF+NEF node.
For external exposure of services related to specific UE(s), the SCEF+NEF resides in the HPLMN. Depending on operator agreements, the SCEF+NEF in the HPLMN may have interface(s) with NF(s) in the VPLMN.
The SCEF+NEF exposes over N33 the same API as the SCEF supports over T8. If CAPIF is not supported, the AF is locally configured with the API termination points for each service. If CAPIF is supported, the AF obtains the service API information from the CAPIF core function via the Availability of service APIs event notification or Service Discover Response as specified in TS 23.222 [64].
The common state information shall be maintained by the combined SCEF+NEF node in order to meet the external interface requirements of the combined node. The common state information includes at least the following data that needs to be common for the SCEF and NEF roles of SCEF+NEF:
- SCEF+NEF ID (must be the same towards the AF).
- SCEF+NEF common IP address and port number.
- Monitoring state for any ongoing monitoring request.
- Configured set of APIs supported by SCEF+ NEF.
- PDN Connection/PDU Session State and NIDD Configuration Information, including Reliable Data Service state information.
- Network Parameter Configuration Information (e.g. Maximum Response Time and Maximum Latency).
The SCEF+NEF need not perform the same procedures for the configuration of monitoring events towards the HSS+UDM twice. For example, if the HSS+UDM is deployed as a combined node, a monitoring event only need to be configured by the SCEF+NEF just once.
The SCEF+NEF may configure monitoring events applicable to both EPC and 5GC using only 5GC procedures towards UDM. In this case, the SCEF+NEF shall indicate that the monitoring event is also applicable to EPC (i.e. the event must be reported both by 5GC and EPC) and may include a SCEF address (i.e. if the event needs to be configured in a serving node in the EPC and the corresponding notification needs to be sent directly to the SCEF). If the HSS and UDM are deployed as separate network entities, UDM shall use HSS services to configure the monitoring event in EPC as defined in TS 23.632 [102]. The UDM shall return an indication to SCEF+NEF of whether the configuration of the monitoring event in EPC was successful. In the case that the UDM reports that the configuration of a monitoring event was not possible in EPC, then the SCEF+NEF may configure the monitoring event using EPC procedures via the HSS as defined in TS 23.682 [36].
NOTE 1: The SCEF+NEF uses only 5GC procedures to configure monitoring events in EPC and 5GC.
NOTE 2: In terms of the CAPIF, the SCEF+NEF is considered a single node. |
fbecc7f0dcf9784c6066646052ab0c0e | 23.501 | 5.17.5.2 Support of interworking for Monitoring Events | |
fbecc7f0dcf9784c6066646052ab0c0e | 23.501 | 5.17.5.2.1 Interworking with N26 interface | In addition to the interworking principles documented in clause 5.17.2.2, the following applies for interworking with N26:
- When UE moves from EPS to 5GS, when the AMF registers in UDM, if no event subscription via UDM is available, the AMF indicates the situation to the UDM and in this case the UDM can decide if the event subscriptions should be provisioned, otherwise if the AMF has event subscription information, after the registration procedure is completed, the AMF may inform the UDM of the currently subscribed events and UDM will do synchronization if needed.
- When UE moves from 5GS to EPS, the MME gets monitoring event configuration from HSS during as part of mobility procedure as specified in clause 4.11.1.3.2 of TS 23.502 [3]. |
fbecc7f0dcf9784c6066646052ab0c0e | 23.501 | 5.17.5.2.2 Interworking without N26 interface | In addition to the interworking principles documented in clause 5.17.2.3, the additional behaviour at EPS to 5GS mobility in clause 5.17.5.2.1 also applies.
When SCEF+NEF performs the procedure of monitoring via the AMF as described in clause 4.15.3.2.4 ("Exposure with bulk subscription") in TS 23.502 [3], if the AMF determines the interworking without N26 interface is supported, during mobility from 5GS to EPS, the AMF shall subscribe on behalf of SCEF+NEF for UDM+HSS notification of MME ID as described in clause 7.1.2 to trigger the SCEF+NEF to configure the monitoring request to the new MME. For single-registration mode, when UE's mobility from 5GS to EPS happens and Serving MME sends Update Location Request to the UDM+HSS, the UDM+HSS provides Serving MME ID to the SCEF+NEF which is the notification endpoint based on the subscription request from AMF. Then the SCEF+NEF performs the procedure of configuring monitoring via the MME for the same Monitoring Events as described in clause 5.6.2.1 of TS 23.682 [36].
When SCEF+NEF performs the procedure of monitoring via the UDM+HSS as described in clause 4.15.3.2.2 of TS 23.502 [3], when UE's mobility between 5GS and EPS happens, the UDM+HSS performs the procedure of configuring monitoring at the MME as described in clause 5.6.1.1 of TS 23.682 [36] and at the AMF as described in clause 4.15.3.2.1 of TS 23.502 [3]. |
fbecc7f0dcf9784c6066646052ab0c0e | 23.501 | 5.17.5.3 Availability or expected level of a service API | A service related with common north-bound API may become unavailable due to UE being served by a CN node not supporting the service. If the availability or expected level of support of a service API associated with a UE changes, for example due to a mobility between 5GC and EPC, the AF shall be made aware of the change.
NOTE 1: If CAPIF is supported and the service APIs become (un)available for the 5GC or EPC network, the AF obtains such information from the CAPIF core function.
If the SCEF+NEF receives the subscription request from the AF for the availability or expected level of support of a service API, the SCEF+NEF subscribes a CN Type Change event for the UE or Group of UEs to the HSS+UDM. If the HSS+UDM receives the subscription for CN Type Change event, the HSS+UDM includes the latest CN type for the UE or Group of UEs in the response for the subscription. If the HSS+UDM detects that the UE switches between being served by the MME and the AMF, the CN Type Change event is triggered and the HSS+UDM notifies the latest CN type for the UE or Group of UEs to the SCEF+NEF. Based on the CN type information, the SCEF+NEF can determine the availability or expected level of support of a given service. The AF will be informed of such information via a subscription/notification service operation. The AF can subscribe for the availability or expected level of support of a service API with report type indicating either One-time report or Continuous report. If there is no CN type information for the UE in the SCEF+NEF, the SCEF+NEF subscribes monitoring event for a new CN Type Change event for the UE or Group of UEs to the HSS+UDM, otherwise, SCEF+NEF determines the CN type locally in the following conditions:
- If the AF subscribes with report type indicating One-time report, the SCEF+NEF may consider the Freshness Timer of the latest CN type information for the UE or Group of UEs. The Freshness Timer is a parameter that is configured based on local SCEF+NEF policy. When a subscription request with One-time report type is received the SCEF+NEF checks if there is the latest CN type information received from the HSS+UDM for the indicated UE ID or External Group ID. If the elapsed time for the CN type information since the last reception is less than the Freshness Timer, then the SCEF+NEF may respond to the AF with the latest CN type information in order to avoid repeated query to HSS+UDM.
- The SCEF+NEF has established a direct connection with MME or AMF or SMF.
When the UE or all members of a Group of UEs are being served by a MME, EPC is determined as CN type. When the UE or all members of a Group of UEs are being served by an AMF, 5GC is determined as CN type. When the UE is registered both in EPC and 5GC, or some members of a Group of UEs are registered in EPC while some members are registered in 5GC, 5GC+EPC is determined as CN type.
NOTE 2: If 5GC+EPC is determined as the CN type serving the UE or the group of UEs, the SCEF+NEF determines that service APIs for both 5GC and EPC are available to the UE or the group of UEs. |
fbecc7f0dcf9784c6066646052ab0c0e | 23.501 | 5.17.6 Void | |
fbecc7f0dcf9784c6066646052ab0c0e | 23.501 | 5.17.7 Configuration Transfer Procedure between NG-RAN and E-UTRAN | |
fbecc7f0dcf9784c6066646052ab0c0e | 23.501 | 5.17.7.1 Architecture Principles for Configuration Transfer between NG-RAN and E-UTRAN | The purpose of the Configuration Transfer between NG-RAN and E-UTRAN is to enable the transfer the RAN configuration information between the gNB and eNodeB via MME and AMF.
In order to make the information transparent for the MME and AMF, the information is included in a transparent container. The source and target RAN node addresses, which allows the Core Network nodes to route the messages. The mechanism depicted in Figure 5.17.7.1-1.
Figure 5.17.7.1-1: Configuration Transfer between gNB and E-UTRAN basic network architecture
The NG-RAN transparent containers are transferred from the source NG-RAN node to the destination E-UTRAN node and vice versa by use of Configuration Transfer messages.
An ENB Configuration Transfer message is used from the E-UTRAN node to the MME over S1 interface as described in TS 36.413 [100], the destination RAN node includes the en-gNB Identifier and may include a TAI associated with the en-gNB. If MME is aware that the en-gNB serves cells which provide access to 5GC, the MME relays the request towards a suitable AMF via inter-system signalling based on a broadcast 5G TAC. An AMF Configuration Transfer message is used from the AMF to the NG-RAN over N2 interface.
A Configuration Transfer message is used by the gNB node to the AMF over N2 interface for the reply and a Configuration Transfer Tunnel message is used to tunnel the transparent container from AMF to MME over the N26 interface. MME relays this reply to the target eNB using a MME CONFIGURATION TRANSFER message. Transport of the RAN containers in E-UTRAN is specified in TS 23.401 [26].
Each Configuration Transfer message carrying the transparent container is routed and relayed independently by the core network node(s). Any relation between messages is transparent for the AMF and MME, i.e. a request/response exchange between applications, for example SON applications, is routed and relayed as two independent messages by the AMF and MME. |
fbecc7f0dcf9784c6066646052ab0c0e | 23.501 | 5.17.7.2 Addressing, routing and relaying | |
fbecc7f0dcf9784c6066646052ab0c0e | 23.501 | 5.17.7.2.1 Addressing | All the Configuration Transfer messages contain the addresses of the source and destination RAN nodes.
An gNB node is addressed by the Target NG-RAN node identifier as described in TS 38.413 [34].
An eNodeB is addressed by the Target eNodeB identifier as described in TS 36.413 [100]. |
fbecc7f0dcf9784c6066646052ab0c0e | 23.501 | 5.17.7.2.2 Routing | The source RAN node sends a message to its core network node including the source and destination addresses.
MME uses the destination address to route the message to the correct AMF via N26 interface. AMF uses the destination address to route the message to the correct MME via N26 interface.
The AMF connected to the destination RAN node decides which RAN node to send the message to, based on the destination address.
The MME connected to the destination RAN node decides which RAN node to send the message to, based on the destination address. |
fbecc7f0dcf9784c6066646052ab0c0e | 23.501 | 5.17.7.2.3 Relaying | The AMF performs relaying between N2 and N26 messages as described in TS 38.413 [34] and TS 29.274 [101].
The MME performs relaying between S1 and N26 message as described in TS 36.413 [100] and TS 29.274 [101]. |
fbecc7f0dcf9784c6066646052ab0c0e | 23.501 | 5.17.8 URSP Provisioning in EPS | When the UE registers in 5GS, the UE includes the Indication of URSP Provisioning Support in EPS in the UE Policy Container carried in Registration Request. On receiving this indication in the UE Policy Container, the PCF will enable the provisioning of the URSP to UE in EPS.
The UE may include the indication of URSP Provisioning Support in EPS in PCO or ePCO to the SMF+PGW-C during:
- Initial Attach with default PDN connection establishment (according to clause 5.3.2.1 of TS 23.401 [26]).
- Any UE requested PDN connectivity request to an additional PDN (according to clause 5.10.2 of TS 23.401 [26].
- Any UE requested bearer resource modification procedure (according to clause 5.4.5 of TS 23.401 [26]).
If the SMF+PGW-C receives the indication of URSP Rule provisioning in EPS from the UE, and it supports URSP provisioning in EPS, it provides the Indication of URSP Provisioning Support in EPS in ePCO in the Create Session Response message to the UE. Then the UE and the SMF+PGW-C considers that the PDN connection is associated with the URSP provisioning in EPS.
The UE may also include the indication of URSP Provisioning Support in EPS in ePCO in the UE Requested PDU Session Establishment procedure (according to clause 4.3.2.2 of TS 23.502 [3] and clause 6.4.1.2 of TS 24.501 [47]). If the SMF+PGW-C receives the indication of URSP Rule provisioning in EPS from the UE and the network supports URSP provisioning in EPS, it provides the Indication of URSP Provisioning Support in EPS in ePCO in the PDU Session Establishment Accept message to the UE, see clause 6.4.1.3 of TS 24.501 [47].
Then the UE and the SMF+PGW-C considers that the PDU Session is associated with the URSP provisioning in EPS. The SMF+PGW-C provides the Indication of URSP Provisioning Support in EPS to the PCF during the SM Policy Association Establishment procedure.
When the UE receives an indication of URSP provisioning support in EPS in the PDN Connectivity Accept message or in PDU Session Establishment Accept and this PDN connection or the corresponding transferred PDN connection is not released, then for any subsequent PDN connectivity requests the UE does not include an indication of URSP Provisioning Support in EPS.
When the UE receives the Indication of URSP Provisioning Support in EPS included in ePCO in the PDN Connectivity Accept message, then the UE initiates the UE requested bearer resource modification with or without QoS update procedure and includes the UE Policy Container ePCO in the Request Bearer Resource Modification message, the UE Policy Container in ePCO will be further forwarded by MME to SMF+PGW-C. When the UE Policy Container ePCO is received by SMF+PGW-C, it forwards transparently the UE Policy Container to PCF for the PDU Session, then the PCF for the PDU Session establishes the UE Policy Association with PCF for the UE. The PCF for the UE generates the corresponding URSP rules in a similar way as it is done in 5GS and sends the URSP rules to UE in the UE Policy Container as described in clause 4.11.0a.5 of TS 23.502 [3]. If the PDN connection which is used for the exchange of UE Policy Container for URSP provisioning is released, the PCF for the PDU Session terminates the SM Policy Association, which then triggers the release of the UE Policy Association with PCF for the UE, see clause 4.11.0a.2a.8 of TS 23.502 [3]. The UE may use the Request Bearer Resource Modification message, over another PDN connection to include the Indication of URSP Provisioning Support in EPS. If the UE receives Indication of URSP Provisioning Support in EPS included in ePCO, then the UE sends a UE requested bearer resource modification with or without QoS update procedure and includes the UE Policy Container ePCO in the Request Bearer Resource Modification message. This triggers the establishment of the UE Policy Association; the UE may receive URSP rules in this PDN connection.
NOTE 1: The UE includes the Indication of URSP Provisioning Support in EPS in PCO or ePCO in the PDN connectivity request according to clause 6.6.1.1 of TS 24.301 [13].
NOTE 2: The MME and Serving-GW can be configured to prioritize the selection of a SMF+PGW-C that support URSP Rule provisioning in EPS.
NOTE 3: When URSP provisioning in EPS is required, the network deployment ensures that the MME supports ePCO.
If the UE does not receive the Indication of URSP Provisioning Support in EPS in ePCO in the PDN Connectivity Accept message, then the UE does not initiate the UE requested bearer resource modification procedure to send the UE Policy Container. The UE may include an indication of URSP Provisioning Support in EPS in a subsequent PDN connectivity request.
The PDN Connection used by UE and SMF+PGW-C to convey UE Policy Container PCO shall be kept when the UE is in the CONNECTED mode. When the UE is attached to EPS, the PCF for the PDU Session retrieves the PCRTs for UE Policy from PCF for the UE and subscribes to the applicable PCRTs to SMF+PGW-C. The PCF for the PDU Session may send the UE Policy Container in the same PDN connection in which the UE Policy Container was received or in any additional PDN connection established with a SMF+PGW-C supporting URSP delivery in EPS.
During EPS to 5GS mobility with N26, the UE Policy Association is terminated by PCF for PDU Session towards the PCF for the UE when it receives the indication of RAT type change from the SMF+PGW-C. The PCF for the PDU Session maintains Indication of URSP Provisioning Support in EPS while the SM Policy Association is established.
During 5GS to EPS mobility with N26, the PCF for the PDU Session determines whether the UE supports URSP delivery in EPS by checking Indication of URSP Provisioning Support in EPS that is locally stored. If the UE and at least one of the SMF+PGW-C for which the UE maintains a PDU session support URSP delivery in EPS the PCF for the PDU session establishes a UE Policy Association with the PCF for the UE. The PCF for the PDU Session discovers the address of PCF for the UE serving the UE by querying BSF. The PCF for the UE recovers the information about the PSI list in the UE and the subscribed PCRTs in 5GS from former UE Policy Association for the UE after receiving the UE Policy Association Establishment request including a UE Policy Container only including an indication about the trigger for the UE Policy Association Establishment ("5GS to EPS mobility").
After the 5GS to EPS mobility, if PCF for the UE needs to provision the URSP to UE, the PCF for the UE sends the UE Policy Container in Npcf_UEPolicyControl_UpdateNotify Request to the PCF for the PDU Session as described in clause 4.11.0a.2a.10 of TS 23.502 [3]. In case there are more than one PDN connections for the UE, then the PCF for the PDU Session selects any of the ongoing PDN Connections via a SMF+PGW-C supporting URSP delivery in EPS for the UE. Then via the selected PDN Connection, the SMF+PGW-C sends the UE Policy Container via ePCO to UE by initiating the PDN GW initiated bearer modification without QoS update procedure as defined in clause 5.4.3 of TS 23.401 [26].
When the UE moves from 5GS to EPS, the PDN connection corresponding to a PDU Session supporting URSP provisioning in EPS is considered as supporting URSP provisioning in EPS. When the UE moves from EPC to 5GS, the PDU Session corresponding to a PDN connection supporting URSP provisioning in EPS is considered as supporting URSP provisioning in EPS.
NOTE 4: At 5GS to EPS mobility with N26, the guard timer in the AMF (as specified in clauses 4.11.1.2.1 and 4.11.1.3.2 of TS 23.502 [3]) ensures that the UE Policy Association remains until the PCF for the UE detects that a UE Policy Association establishment is received from a PCF for the PDU Session indicating 5GS to EPS mobility.
NOTE 5: In the case that the UE is still registered and reachable over non-3GPP access, the PCF for UE can have two UE policy associations for one UE and based on local configuration the PCF can decide to use one of the UE policy associations to update URSP rule to the UE. The UE can receive URSP Rules over any of these two accesses.
NOTE 6: In the case that during the 5GS to EPS mobility there are more than one PCF for the PDU session maintaining PDN connections for the UE every involved PCF for the PDU session can establish a UE Policy Association towards the PCF for the UE. In such case when the PCF for the UE needs to provision the URSP to UE, the PCF for the UE first selects one of those UE Policy Associations for the delivery of URSP rules to the UE in EPS.
The UE can receive ePCO with UE Policy Container over any of the established PDN connections and then response over the same PDN connection.
When the PCF for the UE decides to update the URSPs in the UE via EPS, the PCF for the UE sends the updated URSP rules in UE Policy Container to the PCF for the PDU Session, then the PCF for the PDU Session forwards it to the SMF+PGW-C. The SMF+PGW-C transfers the received UE Policy Container via ePCO to the UE by triggering PDN-GW initiated Bearer without QoS Modification procedure as described in clause 5.4.3 of TS 23.401 [26]. After update the UE policy provided by the PCF, the UE response about the delivery result via ePCO to the network. The SMF+PGW-C transparently forwards the UE response to the PCF for the PDU Session via Update Bearer Response message and then the PCF for the PDU Session forwards it to the PCF for the UE via Npcf_UEPolicyControl_Update Request. If the SMF+PGW-C receives a rejection for Update Bearer Request message (e.g. due to paging failure), the delivery failure result is sent to PCF for the PDU Session and the PCF for the UE. To request to forward the result of delivery of UE policies, "Result of UE Policy Container delivery via EPS" PCRT is applied to the PCF for the PDU Session and the SMF+PGW-C as described in clause 4.11.0a.2a.10 of TS 23.502 [3]. |
fbecc7f0dcf9784c6066646052ab0c0e | 23.501 | 5.18 Network Sharing | |
fbecc7f0dcf9784c6066646052ab0c0e | 23.501 | 5.18.1 General concepts | A network sharing architecture shall allow multiple participating operators to share resources of a single shared network according to agreed allocation schemes. The shared network includes a radio access network. The shared resources include radio resources.
The shared network operator allocates shared resources to the participating operators based on their planned and current needs and according to service level agreements.
In this Release of the specification, the 5G Multi-Operator Core Network (5G MOCN) network sharing architecture, in which only the RAN is shared in 5G System, is supported. The 5G System may also support Indirect Network Sharing deployment between hosting operator (i.e. shared network operator) and participating operator (see clause 6.21 of TS 22.261 [2], Figure 5.18.1-2 and Annex R), in which the RAN is shared. The communication between the shared RAN and the core network of the participating operator is routed through the core network of the hosting operator that connects to the shared RAN.
5G MOCN for 5G System, including UE, RAN and AMF, shall support operators' ability to use more than one PLMN ID (i.e. with same or different country code (MCC) some of which is specified in TS 23.122 [17] and different network codes (MNC)) or combinations of PLMN ID and NID. 5G MOCN supports NG-RAN Sharing with or without multiple Cell Identity broadcast as described in TS 38.300 [27]. Indirect Network Sharing for 5G system, including UE, shared RAN and CP NFs of hosting operator, shall support each participating operator to use more than one PLMN ID.
For Indirect Network Sharing, the shared RAN broadcasts multiple PLMN IDs, including the PLMN ID which represents the hosting operator and the PLMN IDs which represent participating operators. Multiple PLMN IDs are supported by the serving AMF (i.e. the AMF in the core network of the hosting operator). A UE from a participating operator can select the PLMN ID representing the participating operator in the shared RAN area based on existing procedures specified in TS 23.122 [17]. The serving AMF selects core network functions in the PLMN of the participating operator for the UE, based on home routed roaming architecture principle as specified in clause 4.2.4. In addition, the serving AMF selects the SMF of participating operator possibly considering UE location information and also selects a V-SMF in its own network during the PDU session establishment procedure. The serving PLMN ID used in the procedures is determined as follows:
- For interaction between the Network Functions in the hosting operator network and the participating operator network (e.g. AUSF, UDM), the NF of hosting operator network sets the serving PLMN ID to the selected PLMN ID. This includes the cases when the interaction is via another NF (e.g. V-PCF for UE policy, V-SMF) in the hosting operator network.
- For interactions between the Network Functions only in the hosting operator network (e.g. V-PCF for AM policy), the NF of hosting operator network sets the serving PLMN ID to the PLMN ID of the hosting operator.
NOTE 1: In this Release of specification, the Indirect Network Sharing is only applicable for NR with 5GC of hosting operator and 5GC of participating operator. There are maximum of two SMFs (i.e. V-SMF of the hosting operator and H-SMF of the participating operator) controlling a PDU session.
5G MOCN also supports the following sharing scenarios involving non-public networks, i.e.NG-RAN can be shared by any combination of PLMNs, PNI-NPNs (with CAG) and SNPNs (each identified by PLMN ID and NID).
NOTE 2: PNI-NPNs (without CAG) are not explicitly listed above as it does not require additional NG-RAN sharing functionality compared to sharing by one or multiple PLMNs.
In all non-public network sharing scenarios, each Cell Identity as specified in TS 38.331 [28] is associated with one of the following configuration options:
- one or multiple SNPNs;
- one or multiple PNI-NPNs (with CAG); or
- one or multiple PLMNs only.
NOTE 3: This allows the assignment of multiple cell identities to a cell and also allows the cell identities to be independently assigned, i.e. without need for coordination, by the network sharing partners, between PLMNs and/or non-public networks.
NOTE 4: Different PLMN IDs (or combinations of PLMN ID and NID) can also point to the same 5GC. When same 5GC supports multiple SNPNs (identified by PLMN ID and NID), it is up to the operator's policy whether they are used as equivalent SNPNs for a UE.
NOTE 5: There is no standardized mechanism to avoid paging collisions if the same 5G-S-TMSI is allocated to different UEs by different PLMNs or SNPNs of the shared network, as the risk of paging collision is assumed to be very low. If such risk is to be eliminated then PLMNs and SNPNs of the shared network needs to coordinate the value space of the 5G-S-TMSI to differentiate the PLMNs and SNPNs of the shared network.
Figure 5.18.1-1: A 5G Multi-Operator Core Network (5G MOCN) in which multiple CNs are
connected to the same NG-RAN
Figure 5.18.1-2: Indirect Network Sharing in which multiple participating operators' CNs connect to hosting operator's CN to share NR
NOTE 6: Not all interfaces between the hosting operator and the participating operator are depicted in the Figure 5.18.1-2 for simplicity.
NOTE 7: For the sake of clarity, SEPPs are not depicted in the Figure 5.18.1-2.
NOTE 8: In this Release, if the participating operator deploys the Indirect Network Sharing and non-3GPP access (e.g. N3IWF) to 5GC simultaneously, and if the same PLMN ID is used for 3GPP access via Indirect Network Sharing and non 3GPP access (e.g. via N3IWF) to 5GC of HPLMN/participating PLMN, and if there is no N2 interface between the AMF supporting Indirect Network Sharing at the hosting PLMN and the non-3GPP access (e.g. N3IWF) to 5GC of the HPLMN, then the UE cannot be served simultaneously by the 3GPP access via Indirect Network Sharing and non-3GPP access (e.g. via N3IWF) to 5GC of the HPLMN.
To avoid this case, if the participating operator plans to deploy the Indirect Network Sharing and non-3GPP access (e.g. N3IWF) to 5GC simultaneously, it is recommended that one of following measures is used:
1) the different PLMN IDs are used for 3GPP access via Indirect Network Sharing and non 3GPP access (e.g. via N3IWF) to 5GC of HPLMN.
2) the N2 interface is deployed between the AMF(s) of the hosting PLMN supporting Indirect Network Sharing and the non-3GPP access (e.g. N3IWF) to 5GC of the HPLMN.
If none of the two measures in NOTE 8 can be used, the network rejects the UE access via non-3GPP access (e.g. via N3IWF) to 5GC of HPLMN with an existing cause value that will not allow access to 5GC of HPLMN via non-3GPP access (e.g. via N3IWF) when it detects the UE has been registered to its HPLMN ID over 3GPP access via Indirect Network Sharing.
NOTE 9: If none of the two measures in the NOTE 8 can be used and the UE registers the non 3GPP access (e.g. via N3IWF) to 5GC of HPLMN first, then the network can accept the registration to HPLMN ID over 3GPP access via Indirect Network Sharing. How to handle this is up to implementation. |
fbecc7f0dcf9784c6066646052ab0c0e | 23.501 | 5.18.2 Broadcast system information for network sharing | If a shared NG-RAN is configured to indicate available networks (PLMNs and/or SNPNs) for selection by UEs, each cell in the shared radio access network shall in the broadcast system information include available core network operators in the shared network, including the PLMN IDs which represent the participating operators in case of Indirect Network Sharing as specified in clause 5.18.1.
The Broadcast System Information broadcasts a set of PLMN IDs and/or PLMN IDs and NIDs and one or more additional set of parameters per PLMN e.g. cell-ID, Tracking Areas, CAG Identifiers. All 5G System capable UEs that connect to NG-RAN support reception of multiple PLMN IDs and per PLMN specific parameters. All SNPN-enabled UEs support reception of multiple combinations of PLMN ID and NID and SNPN-specific parameters.
The available core network operators (PLMNs and/or SNPNs) shall be the same for all cells of a Tracking Area in a shared NG-RAN network.
UEs not set to operate in SNPN access mode decode the broadcast system information and take the information concerning available PLMN IDs into account in PLMN and cell (re-)selection procedures. UEs set to operate in SNPN access mode decode the broadcast system information and take the information concerning available PLMN IDs and NIDs into account in network and cell (re-)selection procedures. Broadcast system information is specified in TS 38.331 [28] for NR, TS 36.331 [51] for E-UTRA and related UE access stratum idle mode procedures in TS 38.304 [50] for NR and TS 36.304 [52] for E-UTRA.
NOTE: In the case of Indirect Network Sharing, the PLMN ID representing the participating operator which is broadcasted in the Broadcast System Information can be a PLMN ID which is different from HPLMN ID of the participating operator.
5.18.2a PLMN list and SNPN list handling for network sharing
The AMF prepares lists of PLMN IDs or SNPN IDs suitable as target PLMNs or target SNPNs for use at idle mode cell (re)selection and for use at handover and RRC Connection Release with redirection. The AMF:
- provides the UE with the list of PLMNs or list of SNPNs that the UE shall consider as Equivalent to the serving PLMN or the serving SNPN (see TS 23.122 [17]); and
- provides the NG-RAN with a prioritised list of permitted PLMNs or a prioritized list of permitted SNPNs. When prioritising these PLMNs or SNPNs, the AMF may consider the following information: HPLMN of the UE or the subscribed SNPN of the UE, the serving PLMN or the serving SNPN, a preferred target PLMN (e.g. based on last used EPS PLMN) or a preferred target SNPN, or the policies of the operator(s).
For a UE registered in an SNPN, the AMF shall not provide a list of equivalent PLMNs to the UE and shall not provide a list of permitted PLMNs to NG-RAN.
For a UE registered in a PLMN, the AMF shall not provide a list of equivalent SNPNs to the UE and shall not provide a list of permitted SNPNs to NG-RAN. |
fbecc7f0dcf9784c6066646052ab0c0e | 23.501 | 5.18.3 Network selection by the UE | NOTE 1: This clause applies to UEs not operating in SNPN access mode. Network selection for UEs set to operate in SNPN access mode is described in clause 5.30.2.4.
A UE that has a subscription to one of the sharing core network operators shall be able to select this core network operator while within the coverage area of the shared network and to receive subscribed services from that core network operator.
Each cell in shared NG-RAN shall in the broadcast system information include the PLMN IDs concerning available core network operators in the shared network.
When a UE performs an Initial Registration to a network, one of available PLMNs shall be selected to serve the UE. UE uses all the received broadcast PLMN IDs in its PLMN (re)selection processes which is specified in TS 23.122 [17]. UE shall inform the NG-RAN of the selected PLMN so that the NG-RAN can route correctly. The NG-RAN shall inform the core network of the selected PLMN.
NOTE 2: For Indirect Network Sharing as specified in clause 5.18.1, the broadcast PLMN IDs include the PLMN IDs representing participating operators and a UE from a participating operator selects a PLMN ID representing the participating operator as specified in TS 23.122 [17].
NOTE 3: The serving AMF of hosting operator's network acquires the relevant UE subscription data from the participating operator's network during the Registration procedure as described in clause 4.2.2.2.2 of TS 23.502 [3].
In the case of Indirect Network Sharing, in this Release of the specifications, if hosting operator's network identifies that a UE is not from a participating operator when the UE selects a PLMN ID representing the participating operator, then the hosting operator network rejects the UE with an existing cause value during the registration procedure.
As per any network, after Initial Registration to the shared network and while remaining served by the shared network, the network selection procedures specified in TS 23.122 [17] may cause the UE to perform a reselection of another available PLMN.
UE uses all of the received broadcast PLMN IDs in its cell and PLMN (re)selection processes. |
fbecc7f0dcf9784c6066646052ab0c0e | 23.501 | 5.18.4 Network selection by the network | The NG-RAN uses the selected PLMN/SNPN (provided by the UE at RRC establishment, or, provided by the AMF/source NG-RAN at N2/Xn handover) to select target cells for future handovers (and radio resources in general) appropriately. The network should not move the UE to another available PLMN/SNPN, e.g. by handover, as long as the selected PLMN/SNPN is available to serve the UE's location.
In the case of handover or network controlled release to a PLMN in a shared network:
- When multiple PLMN IDs are broadcasted in a cell selected by NG-RAN, NG-RAN shall select a target PLMN, taking into account the prioritized list of PLMN IDs provided via Mobility Restriction List from AMF.
- For Xn based HO procedure, Source NG-RAN indicates the selected PLMN ID to the target NG-RAN, see TS 38.300 [27].
- For N2 based HO procedure, the NG-RAN indicates a selected PLMN ID to the AMF as part of the TAI sent in the HO required message. Source AMF uses the TAI information supplied by the source NG-RAN to select the target AMF/MME. The source AMF should forward the selected PLMN ID to the target AMF/MME. The target AMF/MME indicates the selected PLMN ID to the target NG-RAN/eNB so that the target NG-RAN/eNB can select target cells for future handover appropriately.
- For RRC connection release with redirection to E-UTRAN procedure, NG-RAN decides the target network by using PLMN information as defined in the first bullet.
A change in serving PLMN is indicated to the UE as part of the UE registration with the selected network via 5G-GUTI in 5GS.
In the case of handover or network controlled release to an SNPN in a shared network, the following applies:
- When multiple SNPN IDs are broadcasted in a cell selected by NG-RAN, NG-RAN shall select a target SNPN, taking into account the prioritized list of SNPN IDs provided via Mobility Restriction List from AMF.
- For Xn based HO procedure, Source NG-RAN indicates the selected SNPN ID to the target NG-RAN, see TS 38.300 [27].
- For N2 based HO procedure, the NG-RAN indicates a selected SNPN ID to the AMF together with the TAI sent in the HO required message. Source AMF uses the selected SNPN ID together with the TAI information supplied by the source NG-RAN to select the target AMF. The source AMF should forward the selected SNPN ID to the target AMF. The target AMF indicates the selected SNPN ID to the target NG-RAN so that the target NG-RAN can select target cells for future handover appropriately.
A change in serving SNPN is indicated to the UE as part of the UE registration with the selected network. |
fbecc7f0dcf9784c6066646052ab0c0e | 23.501 | 5.18.5 Network Sharing and Network Slicing | As defined in clause 5.15.1, a Network Slice is defined within a PLMN or SNPN. Network sharing is performed among different PLMNs and/or SNPNs.
In the case of 5G MOCN, each PLMN or SNPN sharing the NG-RAN defines and supports its PLMN- or SNPN- specific set of slices that are supported by the common NG-RAN.
In the case of Indirect Network Sharing, the differences of Network Slicing handling is performed as the follows:
- If, in an Indirect Network Sharing area, the selected PLMN ID which is broadcasted by shared RAN is the HPLMN ID or EHPLMN ID (which is not derived from UE's SUPI) of the UE, from the network side, the network slicing handling shall be performed as defined in clause 5.15.4 and clause 5.15.6. The changes compared to baseline operations are as follows:
- During Registration procedure, the UE includes S-NSSAI(s) of participating operator in the NAS Registration Request messages as described in TS 24.501 [47]. After receiving the Registration Request including the above information, the serving AMF determines the corresponding S-NSSAIs of hosting operator based on the Subscribed S-NSSAIs and slice mapping information from the HPLMN S-NSSAI(s) to the S-NSSAI(s) of the hosting operator locally or the slice mapping information received from the NSSF of hosting operator's network. The serving AMF further checks whether the corresponding S-NSSAIs of hosting operator are supported by the shared RAN. If none of the corresponding S-NSSAIs of hosting operator is supported, the serving AMF rejects the Registration Request. Otherwise the serving AMF continues the Registration procedure and sends to the UE Allowed NSSAI, Partially Allowed NSSAI, Rejected S-NSSAIs and Configured NSSAI only containing the S-NSSAI(s) of participating operator, i.e. HPLMN or EHPLMN S-NSSAI(s) without mapping information. The serving AMF creates a list of S-NSSAIs of the hosting operator corresponding to the S-NSSAIs included in the Allowed NSSAI sent to the UE and includes these S-NSSAIs in the Allowed NSSAI in the signalling to the shared RAN and to the PCF during the AM policy association establishment. The UE Context in the serving AMF shall retain the slice mapping information. The above AMF functionality is applied also during UE Configuration Update procedure.
NOTE 1: In the case of Indirect Network Sharing, to support EHPLMN ID broadcast option, as deployment requirement, the S-NSSAI value supported by the EHPLMN is the same as the HPLMN S-NSSAI value.
- During PDU Session Establishment procedure, the UE includes the S-NSSAI(s) of participating operator in the PDU Session Establishment Request as described in TS 24.501 [47]. After receiving the S-NSSAI of participating operator from UE, the serving AMF uses the slice mapping information stored in the UE Context to determine the S-NSSAI of hosting operator (i.e. VPLMN S-NSSAI) to be used as "the S-NSSAI belonging to the Allowed NSSAI" as defined in clause 5.15.6. The serving AMF sends the serving PLMN ID (i.e. the selected PLMN ID representing the participating operator), VPLMN S-NSSAI (i.e. the corresponding S-NSSAI of the hosting operator) and HPLMN S-NSSAI to the V-SMF. The V-SMF can identify the Indirect Network Sharing case based on the serving PLMN ID and then it sends the HPLMN S-NSSAI in the PDU Session Establishment Accept message to the UE and uses the VPLMN S-NSSAI for the user plane handling (NG-RAN, V-UPF) in hosting operator's network.
NOTE 2: If NSSF is involved in the above cases, the AMF will send HPLMN S-NSSAI(s) to the NSSF to retrieve the slice mapping information from the HPLMN S-NSSAI(s) to the VPLMN S-NSSAI(s).
NOTE 3: In the case of Indirect Network Sharing, HPLMN/EHPLMN S-NSSAI is corresponding to S-NSSAI of participating operator and VPLMN S-NSSAI is corresponding to S-NSSAI of hosting operator.
NOTE 4: In the case of Indirect Network Sharing, the shared RAN only needs to support the S-NSSAI of hosting operator and select the serving AMF as described in clause 5.15.5.2.1. Then, the shared RAN does not need to be provisioned with the slice mapping information. The serving AMF and the V-SMF in the hosting operator network can be configured to know whether the serving PLMN ID (i.e. the selected PLMN ID) is HPLMN ID/EHPLMN ID or not. The serving AMF and V-SMF with the above functionality will be selected in this case.
- For other cases, i.e. the broadcast PLMN ID which represents the participating operator is different from HPLMN ID and EHPLMN IDs of UE, the network slicing handling including UE and network shall comply with the roaming case as defined in clause 5.15.6. The applied S-NSSAI values for the PLMN ID selected by UE (i.e. the broadcast PLMN ID) are the S-NSSAI values of hosting operator.
NOTE 5: Which broadcast option to be used and whether to broadcast one or more PLMN ID(s) for each participating operator is determined based on the agreement between the hosting operator and the participating operator. |
fbecc7f0dcf9784c6066646052ab0c0e | 23.501 | 5.19 Control Plane Load Control, Congestion and Overload Control | |
fbecc7f0dcf9784c6066646052ab0c0e | 23.501 | 5.19.1 General | In order to ensure that the network functions within 5G System are operating under nominal capacity for providing connectivity and necessary services to the UE. Thus, it supports various measures to guard itself under various operating conditions (e.g. peak operating hour, extreme situations). It includes support for load (re-)balancing, overload control and NAS level congestion control. A 5GC NF is considered to be in overload when it is operating over its nominal capacity resulting in diminished performance (including impacts to handling of incoming and outgoing traffic).
AMF and SMFs may, based on operator's policy/configuration and/or upon detection of certain local events, e.g. massive UEs signalling from a UE group, or based on analytics about expected NF load for selected NFs in the network, subscribe at the NWDAF for Signalling Storm Analytics as specified in TS 23.288 [86] to obtain information related to signalling storms caused by excessive UE or NF signalling and may decide to take mitigations actions as described in the present clause 5.19 based on the received analytics information.
In addition, 5GC NFs can consider signalling storm analytics provided by NWDAF as defined in clause 6.22.3 of TS 23.288 [86], to determine the control plane load Control, congestion and overload control, detail example actions of NFs are described in Table 6.22.3-3 in clause 6.22.3 of TS 23.288 [86]. |
fbecc7f0dcf9784c6066646052ab0c0e | 23.501 | 5.19.2 TNLA Load Balancing and TNLA Load Re-Balancing | AMF can support load balancing and re-balancing of TNL associations between 5G-AN and AMF by using mechanisms specified in clause 5.21.1. |
fbecc7f0dcf9784c6066646052ab0c0e | 23.501 | 5.19.3 AMF Load Balancing | The AMF Load Balancing functionality permits UEs that are entering into an AMF Region/AMF Set to be directed to an appropriate AMF in a manner that achieves load balancing between AMFs. This is achieved by setting a Weight Factor for each AMF, such that the probability of the 5G-AN selecting an AMF is proportional to Weight Factor of the AMF. The Weight Factor is typically set according to the capacity of an AMF node relative to other AMF nodes. The Weight Factor is sent from the AMF to the 5G-AN via NGAP messages (see TS 38.413 [34]).
NOTE 1: An operator may decide to change the Weight Factor after the establishment of NGAP connectivity as a result of changes in the AMF capacities. e.g. a newly installed AMF may be given a very much higher Weight Factor for an initial period of time making it faster to increase its load.
NOTE 2: It is intended that the Weight Factor is NOT changed frequently. e.g. in a mature network, changes on a monthly basis could be anticipated, e.g. due to the addition of 5G-AN or 5GC nodes.
NOTE 3: Weight Factors for AMF Load Balancing are associated with AMF Names.
Load balancing by 5G-AN node is only performed between AMFs that belong to the same AMF set, i.e. AMFs with the same PLMN, AMF Region ID and AMF Set ID value.
The 5G-AN node may have their Load Balancing parameters adjusted (e.g. the Weight Factor is set to zero if all subscribers are to be removed from the AMF, which will route new entrants to other AMFs within an AMF Set). |
fbecc7f0dcf9784c6066646052ab0c0e | 23.501 | 5.19.4 AMF Load Re-Balancing | The AMF load re-balancing functionality permits cross-section of its subscribers that are registered on an AMF (within an AMF Set) to be moved to another AMF within the same AMF set with minimal impacts on the network and end users. AMF may request some or all of the 5G-AN node(s) to redirect a cross-section of UE(s) returning from CM-IDLE state to be redirected to another AMF within the same AMF set, if the 5G-AN is configured to support this. The AMF may request some or all of the 5G-AN node(s) to redirect the UEs served by one of its GUAMI(s) to a specific target AMF within the same AMF set or to any different AMF within the same AMF set.
When indicating a specific target AMF, the AMF should ensure that the load re-balancing will not cause overload in the target AMF.
NOTE: This requirement can be fulfilled by the AMF itself or by the OAM.
For UE(s) in CM-IDLE state, when UE subsequently returns from CM-IDLE state and the 5G-AN receives an initial NAS message with a 5G S-TMSI or GUAMI pointing to an AMF that requested for redirection, the 5G-AN should select the specific target AMF (provided by the original AMF) or a different AMF from the same AMF set and forward the initial NAS message.
For UE(s) in CONNECTED mode, similar mechanisms for AMF Management can be used to move the UE to another AMF in the same AMF set as described in clause 5.21.2, except that the old AMF deregisters itself from NRF.
The newly selected/target AMF (which is now the serving AMF) will re-assign the GUTI (using its own GUAMI(s)) to the UE(s). It is not expected that the 5G-AN node rejects any request or enables access control restriction when it receives a request for redirection for load control from the connected AMF(s).
When the AMF wants to stop redirection, the AMF can indicate that it can serve all UE(s) in CM-IDLE state to stop the redirection.
NOTE 1: An example use for the AMF load re-balancing functionality is for the AMF to pro-actively re-balance its load prior to reaching overload i.e. to prevent overload situation.
NOTE 2: Typically, AMF Load Re-Balancing is not needed when the AMF becomes overloaded because the Load Balancing function should have ensured that the other AMFs within the AMF Set are similarly overloaded. |
fbecc7f0dcf9784c6066646052ab0c0e | 23.501 | 5.19.5 AMF Control Of Overload | |
fbecc7f0dcf9784c6066646052ab0c0e | 23.501 | 5.19.5.1 General | The AMF shall contain mechanisms for avoiding and handling overload situations. This includes the following measures:
- N2 overload control that could result in RRC reject, RRC Connection Release and unified access barring.
- NAS congestion control. |
fbecc7f0dcf9784c6066646052ab0c0e | 23.501 | 5.19.5.2 AMF Overload Control | Under unusual circumstances, if AMF has reached overload situation or there is high probability that signalling storm will happen based on received analytics of signalling storm caused by NF(s) abnormal signalling or set of UEs in specific group, slice or priority as described in clause 6.22 in TS 23.288 [86], the AMF activates NAS level congestion control as specified in Clause 5.19.7 and AMF restricts the load that the 5G-AN node(s) are generating, if the 5G-AN is configured to support overload control. N2 overload control can be achieved by the AMF invoking the N2 overload procedure (see TS 38.300 [27] and TS 38.413 [34]) to all or to a proportion of the 5G-AN nodes with which the AMF has N2 connections. The AMF may include the S-NSSAI(s) in NGAP OVERLOAD START message sent to 5G-AN node(s) to indicate the Network Slice(s) with which NAS signalling is to be restricted. To reflect the amount of load that the AMF wishes to reduce, the AMF can adjust the proportion of 5G-AN nodes which are sent NGAP OVERLOAD START message and the content of the overload start procedure.
When NGAP OVERLOAD START is sent by multiple AMFs or from the same AMF set in the same PLMN towards the 5G-AN, it should be ensured that the signalling load is evenly distributed within the PLMN and within each AMF set.
A 5G-AN node supports restricting of 5G-AN signalling connection when a signalling connection establishment are attempted by certain UEs (which are registered or attempting to register with the 5GC), as specified in TS 38.331 [28] and TS 36.331 [51]. Additionally, a 5G-AN node provides support for the barring of UEs as described in TS 22.261 [2]. These mechanisms are further specified in TS 38.331 [28] and TS 36.331 [51]. For 3GPP Access Type, the signalling connection establishment attempt includes a RRC Connection Resume procedure from RRC_INACTIVE.
By sending the NGAP OVERLOAD START message, the AMF can request the 5G-AN node to apply the following behaviour for UEs that the AMF is serving:
a) Restrict 5G-AN signalling connection requests that are not for emergency, not for exception reporting and not for high priority mobile originated services; or
b) Restrict 5G-AN signalling connection requests for uplink NAS signalling transmission to that AMF;
c) Restrict 5G-AN signalling connection requests where the Requested NSSAI at AS layer only includes the indicated S-NSSAI(s) in the NGAP OVERLOAD START message. This applies also to RRC_INACTIVE Connection Resume procedure where the Allowed NSSAI in the stored UE context in the RAN only includes S-NSSAIs included in the NGAP OVERLOAD START.
d) only permit 5G-AN signalling connection requests for emergency sessions and mobile terminated services for that AMF; or
e) only permit 5G-AN signalling connection requests for high priority sessions, exception reporting and mobile terminated services for that AMF;
The above applies for RRC Connection Establishment procedure and RRC Connection Resume procedures over 3GPP access, as well as for the UE-N3IWF connection establishment over untrusted Non-3GPP access and for the UE-TNGF connection establishment over trusted Non-3GPP access.
The AMF can provide a value that indicates the percentage of connection requests to be restricted in the NGAP OVERLOAD START and the 5G-AN node may consider this value for congestion control.
When restricting a 5G-AN signalling connection, the 5G-AN indicates to the UE an appropriate wait timer that limits further 5G-AN signalling connection requests until the wait timer expires.
During an overload situation, the AMF should attempt to maintain support for emergency services and for MPS.
When the AMF is recovering, the AMF can either:
- send a NGAP OVERLOAD START message with a new percentage value that permits more connection requests to be successful, or
- send a NGAP OVERLOAD STOP message.
to the same 5G-AN node(s) the NGAP OVERLOAD START was previously sent. |
fbecc7f0dcf9784c6066646052ab0c0e | 23.501 | 5.19.6 SMF Overload Control | The SMF shall contain mechanisms for avoiding and handling overload situations. This can include the following measures:
- SMF overload control that could result in rejections of NAS requests.
The SMF overload control may be activated by SMF due to congestion situation at SMF e.g. configuration, by a restart or recovery condition of a UPF, or by a partial failure or recovery of a UPF for a particular UPF(s).
Under unusual circumstances, if the SMF has reached overload situation or there is high probability that signalling storm will happen based on received analytics of signalling storm caused by NF(s) abnormal signalling or set of UEs in specific group, slice or priority as described in clause 6.22 of TS 23.288 [86], the SMF activates NAS level congestion control as specified in clause 5.19.7. The SMF may restrict the load that the AMF(s) are generating, if the AMF is configured to enable the overload restriction. |
fbecc7f0dcf9784c6066646052ab0c0e | 23.501 | 5.19.7 NAS level congestion control | |
fbecc7f0dcf9784c6066646052ab0c0e | 23.501 | 5.19.7.1 General | NAS level congestion control may be applied in general (i.e. for all NAS messages), per DNN, per S-NSSAI, per DNN and S-NSSAI, or for a specific group of UEs.
NAS level congestion control is achieved by providing the UE a back-off time. To avoid that large amounts of UEs initiate deferred requests (almost) simultaneously, the 5GC should select each back-off time value so that the deferred requests are not synchronized. When the UE receives a back-off time, the UE shall not initiate any NAS signalling with regards to the applied congestion control until the back-off timer expires or the UE receives a mobile terminated request from the network, or the UE initiates signalling for emergency services or high priority access.
AMFs and SMFs may apply NAS level congestion control, but should not apply NAS level congestion control for procedures not subject to congestion control. |
fbecc7f0dcf9784c6066646052ab0c0e | 23.501 | 5.19.7.2 General NAS level congestion control | This clause only applies to NAS Mobility Management congestion control.
Under general overload conditions the AMF may reject NAS messages from UEs using any 5G-AN. When a NAS request is rejected, a Mobility Management back-off time may be sent by the AMF to the UE. While the Mobility Management back-off timer is running, the UE shall not initiate any NAS request except for Deregistration procedure and procedures not subject to congestion control (e.g. high priority access, emergency services, informing the network of UE unavailability period due to NR satellite access discontinuous coverage) and mobile terminated services. After any such Deregistration procedure, the back-off timer continues to run. While the Mobility Management back-off timer is running, the UE is allowed to perform Mobility Registration Update if the UE is already in CM-CONNECTED state. If the UE receives a paging request or a NAS notification message from the AMF while the Mobility Management back off timer is running, the UE shall stop the Mobility Management back-off timer and initiate the Service Request procedure or the Mobility Registration Update procedure over 3GPP access and/or non-3GPP access as applicable. Over non-3GPP access, if the UE is in CM-IDLE state when the back-off timer is stopped, it shall initiate the UE-triggered Service Request procedure as soon as it switches back to CM-CONNECTED state.
In order to allow the UE to report the PS Data Off status change in PDU Session Modification Request message, the UE behaves as follows while keeping the NAS MM back-off timer running in the UE:
- When the UE is in CM-IDLE state and has not moved out of the Registration Area, the UE is allowed to send a Service Request message with an indication that the message is exempted from NAS congestion control. When the UE is in CM-IDLE mode and has moved out of the Registration Area, the UE is allowed to send a Mobility Registration Update request message, with a Follow-on request and with an indication that the message is exempted from NAS congestion control.
- When the UE is in CM-CONNECTED state, the UE sends a PDU Session Modification Request with PS Data Off status change carried in UL NAS Transport message with an indication that the message is exempted from NAS congestion control.
When the NAS MM congestion control is activated at AMF, if the UE indicates that the NAS MM message is exempted from NAS congestion control, the AMF shall not reject the NAS MM message and shall forward the NAS SM message to the corresponding SMF with an indication that the NAS SM message was indicated to be exempted from NAS congestion control. The SMF ensures that the NAS SM message is not subject to congestion control otherwise the SMF rejects the message, e.g. the SMF shall reject PDU Session Modification received if it is not for Data Off status reporting.
The Mobility Management back-off timer shall not impact Cell/RAT/Access Type and PLMN change. Cell/RAT/TA/Access Type change does not stop the Mobility Management back-off timer. The Mobility Management back-off timer shall not be a trigger for PLMN reselection or SNPN reselection. The back-off timer is stopped as defined in TS 24.501 [47] when a new PLMN or SNPN, that is not an equivalent PLMN or is not an equivalent SNPN, is accessed.
To avoid that large amounts of UEs initiate deferred requests (almost) simultaneously, the AMF should select the Mobility Management back-off timer value so that the deferred requests are not synchronized. If NWDAF is deployed, the AMF may make use of UE signalling analytics provided by NWDAF, as defined in clause 6.22.3 of TS 23.288 [86], to determine the Mobility Management back-off timer value provided to UEs.
If the UE required to report 5GSM Core Network Capability change, or the Always-on PDU Session Requested indication while the NAS MM congestion control timer was running and was unable to initiate MM signalling, the UE defers the related MM signalling until the MM congestion control timer expires and initiates after the expiry of the timer.
In the case of a UE with scheduled communication pattern, the AMF may consider the UE's communication pattern while selecting a value for the Mobility Management back-off timer so that the UE does not miss its only scheduled communication window.
The AMF should not reject Registration Request message for Mobility Registration Update that are performed when the UE is already in CM-CONNECTED state or when UE is indicating unavailability period due to NR satellite access discontinuous coverage.
The AMF may reject the Service Request message and a UL NAS Transfer with a Mobility Management back-off time when the UE is already in CM-CONNECTED state. If UE receives a DL NAS Transfer message from the AMF while the Mobility Management back off timer is running, the UE shall stop the Mobility Management back-off timer.
For CM-IDLE state mobility, the AMF may reject Registration Request messages for Mobility Registration Update by including a Mobility Management back off time value in the Registration Reject message.
If UE registered in the same PLMN for 3GPP access and non-3GPP access and receives a Mobility Management back-off time from the AMF, the back-off time (and corresponding start and stop) is applied equally to both 3GPP access and non-3GPP access. If UE registered in different PLMNs for 3GPP access and non-3GPP access respectively and receives a Mobility Management back-off time, the back-off time is only applied to the PLMN that provides the time to the UE.
If the AMF rejects Registration Request messages or Service Request with a Mobility Management back-off time which is larger than the sum of the UE's Periodic Registration Update timer and the Implicit Deregistration timer, the AMF should adjust the mobile reachable timer and/or Implicit Deregistration timer such that the AMF does not implicitly deregister the UE while the Mobility Management back-off timer is running.
NOTE: This is to minimize signalling after the Mobility Management back-off timer expires.
If the AMF deregisters the UE with an indication of re-registration required, the UE behaviour for handling the back-off timer(s) is as specified in TS 24.501 [47]. |
fbecc7f0dcf9784c6066646052ab0c0e | 23.501 | 5.19.7.3 DNN based congestion control | DNN based congestion control is designed for the purpose of avoiding and handling of NAS SM signalling congestion for the UEs with a back-off timer associated with or without a DNN regardless of the presence of an S-NSSAI. Both UE and 5GC shall support the functionality to enable DNN based congestion control.
SMFs may apply DNN based congestion control towards the UE by rejecting PDU Session Establishment Request message, or PDU Session Modification Request message except for those sent for the purpose of reporting 3GPP PS Data Off status change for a specific DNN with a running back-off timer. The SMF may release PDU Sessions belonging to a congested DNN by sending a PDU Session Release Command message towards the UE with a DNN back-off timer. If a DNN back-off time is set in the PDU Session Release Command message, the cause value of "reactivation requested" shall not be set.
If NWDAF is deployed, the SMF may make use of Session Management Congestion Control Experience analytics provided by NWDAF, as defined in clause 6.12 of TS 23.288 [86], to determine back-off timer provided to UEs.
NOTE: For example, the SMF can apply a short back-off timer to the UEs in the list of high-experienced UEs while the SMF can apply a long back-off timer to the UEs in the list of low-experienced UEs.
When DNN based congestion control is activated at AMF e.g. configured by OAM, the AMF provides a NAS Transport Error message for the NAS Transport message carrying an SM message and in the NAS Transport Error message it includes a DNN back-off timer.
The UE associates the received back-off time with the DNN (i.e. no DNN, DNN only) which the UE included in the uplink NAS MM message carrying the corresponding NAS SM request message.
The UE associates the received back-off time with the DNN (i.e. no DNN, DNN only) in any PLMN unless the DNN associated with the back-off timer is an LADN DNN in which case the UE only associates it to the PLMN in which the back-off time was received.
The UE behaves as follows when the DNN back-off timer is running:
- If a DNN is associated with the back-off timer, the UE shall not initiate any Session Management procedures for the congested DNN. The UE may initiate Session Management procedures for other DNNs. The UE shall not initiate any Session Management procedure for the corresponding APN when UE moves to EPS. The UE may initiate Session Management procedures for other APNs when the UE moves to EPS;
- If no DNN is associated with the back-off timer, the UE may only initiate Session Management requests of any PDU Session Type for a specific DNN;
- Upon Cell/TA/PLMN/RAT change, change of untrusted non-3GPP access network or change of Access Type, the UE shall not stop the back-off timer;
- The UE is allowed to initiate the Session Management procedures for high priority access and emergency services;
- The UE is allowed to initiate the Session Management procedure for reporting Data Off status change to the network;
- If the UE receives a network initiated Session Management message other than PDU Session Release Command for the congested DNN associated to a running back-off timer, the UE shall stop the back-off timer and respond to the 5GC;
- If the UE receives a PDU Session Release Command message for the congested DNN, it shall stop the back-off timer unless it receives a new back-off time from SMF;
- The UE is allowed to initiate PDU Session Release procedure (i.e. sending PDU Session Release Request message). The UE shall not stop the back-off timer when the related PDU Session is released;
- The list above is not an exhaustive list, i.e. more details of the above actions and further conditions, if any, are specified in TS 24.501 [47].
If UE initiates one of the Session Management procedures that are exempted from NAS congestion control, the UE indicates that the carried NAS SM message is exempted from NAS congestion control in the UL NAS Transport message as described in TS 24.501 [47]. When the DNN based congestion control is activated at AMF, if the UE indicates that the NAS SM message in the UL NAS Transport message is exempted from NAS congestion control, the AMF shall not apply DNN based congestion control on the UL NAS Transport message and shall forward the NAS SM message to the corresponding SMF with an indication that the message was received with exemption indication. The SMF evaluates whether the NAS SM message is allowed to be exempted from DNN based congestion control. If it is not, the SMF rejects the message, e.g. the SMF shall reject PDU Session Modification received if it is not for Data Off status reporting).
The UE shall maintain a separate back-off timer for each DNN that the UE may use.
To avoid that large amounts of UEs initiate deferred requests (almost) simultaneously, the 5GC should select the back-off timer value so that deferred requests are not synchronized.
If the UE required to report 5GSM Core Network Capability change, or the Always-on PDU Session Requested indication while DNN based congestion control was running and was unable to initiate SM signalling, the UE defers the related SM signalling until the DNN based congestion control timer expires and initiates the necessary SM signalling after the expiry of the timer.
The DNN based Session Management congestion control is applicable to the NAS SM signalling initiated from the UE in the Control Plane. The Session Management congestion control does not prevent the UE from sending and receiving data or initiating Service Request procedures for activating User Plane connection towards the DNN(s) that are under Session Management congestion control. |
fbecc7f0dcf9784c6066646052ab0c0e | 23.501 | 5.19.7.4 S-NSSAI based congestion control | S-NSSAI based congestion control is designed for the purpose of avoiding and handling of NAS signalling congestion for the UEs with back-off timer associated with or without an S-NSSAI regardless of the presence of a DNN.
The UE associates the received back-off time with the S-NSSAI and DNN (i.e. no S-NSSAI and no DNN, no S-NSSAI, S-NSSAI only, an S-NSSAI and a DNN) which was included in the uplink NAS MM message carrying the corresponding NAS SM request message for the PLMN which is under congestion.
S-NSSAI based congestion control is applied as follows:
- If an S-NSSAI is determined as congested, then the SMF may apply S-NSSAI based congestion control towards the UE for SM requests except for those sent for the purpose of reporting 3GPP PS Data Off status change for a specific S-NSSAI and provides a back-off time and an indication of HPLMN congestion;
- If the UE receives an S-NSSAI based back-off time without an indication of HPLMN congestion, the UE shall apply the S-NSSAI back-off timer only in the PLMN in which the back-off time was received. If the UE receives S-NSSAI based back-off time with an indication of HPLMN congestion, the UE shall apply the S-NSSAI based back-off timer in the PLMN in which the back-off time was received and in any other PLMN;
- The SMF may release PDU Sessions belonging to a congested S-NSSAI by sending a PDU Session Release Request message towards the UE with a back-off time associated either to the S-NSSAI only (i.e. with no specific DNN) or a combination of the S-NSSAI and a specific DNN. If NWDAF is deployed, the SMF may make use of Session Management Congestion Control Experience analytics provided by NWDAF, as defined in clause 6.12 of TS 23.288 [86], to determine back-off timer provided to UEs;
NOTE: For example, the SMF can apply a short back-off timer to the UEs in the list of high-experienced UEs while the SMF can apply a long back-off timer to the UEs in the list of low-experienced UEs.
- If S-NSSAI based congestion control is activated at AMF e.g. configured by OAM and an S-NSSAI is determined as congested, then the AMF applies S-NSSAI based congestion control towards the UE for UE-initiated Session Management requests. In this case, the AMF provides a NAS Transport Error message for the NAS Transport message carrying the SM message and in the NAS Transport Error message it includes a back-off timer; If NWDAF is deployed, the AMF may determine that S-NSSAI is congested based on the network slice load level analytics defined in TS 23.288 [86].
- The UE behaves as follows in the PLMN where the S-NSSAI based congestion control applies when the back-off timer is running:
- If the back-off timer was associated with an S-NSSAI only (i.e. not associated with an S-NSSAI and a DNN), the UE shall not initiate any Session Management procedures for the congested S-NSSAI;
- If the back-off timer was associated with an S-NSSAI and a DNN, then the UE shall not initiate any Session Management procedures for that combination of S-NSSAI and DNN;
- If the UE receives a network-initiated Session Management message other than PDU Session Release Command for the congested S-NSSAI, the UE shall stop this back-off timer and respond to the 5GC;
- If the UE receives a PDU Session Release Command message for the congested S-NSSAI, it shall stop the back-off timer unless it receives a new back-off time from SMF;
- Upon Cell/TA/PLMN/RAT change, change of untrusted non-3GPP access network or change of Access Type, the UE shall not stop the back-off timer for any S-NSSAI or any combination of S-NSSAI and DNN;
- The UE is allowed to initiate the Session Management procedures for high priority access and emergency services for the S-NSSAI;
- The UE is allowed to initiate the Session Management procedure for reporting Data Off status change for the S-NSSAI or the combination of S-NSSAI and DNN.
- If the back-off timer is not associated to any S-NSSAI, the UE may only initiate Session Management procedures for specific S-NSSAI;
- If the back-off timer is not associated to any S-NSSAI and DNN, the UE may only initiate Session Management procedures for specific S-NSSAI and DNN;
- The UE is allowed to initiate PDU Session Release procedure (e.g. sending PDU Session Release Request message). The UE shall not stop the back-off timer when the related PDU Session is released;
- The list above is not an exhaustive list, i.e. more details of the above actions and further conditions, if any, are specified in TS 24.501 [47].
The UE shall maintain a separate back-off timer for each S-NSSAI and for each combination of S-NSSAI and DNN that the UE may use.
If UE initiates one of the Session Management procedure that are exempt from NAS congestion control, the UE indicates that the carried NAS SM message is exempted from NAS congestion control in the UL NAS Transport message as described in TS 24.501 [47]. When the S-NSSAI based congestion control is activated at AMF, if the UE indicates that the NAS SM message in the UL NAS Transport message is exempted from NAS congestion control, the AMF shall not apply S-NSSAI based congestion control on the UL NAS Transport message and shall forward the NAS SM message to the corresponding SMF with an indication that the message was received with exemption indication. The SMF evaluates whether that the NAS SM message is allowed to be exempted from S-NSSAI based congestion control. If it is not, the SMF rejects the message, e.g. the SMF shall reject PDU Session Modification received if it is not for Data Off status reporting.
The back-off timer associated with an S-NSSAI or a combination of an S-NSSAI and a DNN shall only apply to congestion control for Session Management procedures when UE is in 5GS.
To avoid that large amounts of UEs initiate deferred requests (almost) simultaneously, the 5GC should select the value of the back-off timer for the S-NSSAI based congestion control so that deferred requests are not synchronized.
If the UE required to report 5GSM Core Network Capability change, or the Always-on PDU Session Requested indication while S-NSSAI based congestion control timer was running and was unable to initiate SM signalling, the UE defers the related SM signalling until the S-NSSAI based congestion control timer expires and initiates the necessary SM signalling after the expiry of the timer.
The S-NSSAI based congestion control does not prevent the UE from sending and receiving data or initiating Service Request procedure for activating User Plane connection for a PDU Session associated to the S-NSSAI that is under the congestion control. |
fbecc7f0dcf9784c6066646052ab0c0e | 23.501 | 5.19.7.5 Group specific NAS level congestion control | The group specific NAS level congestion control applies to a specific group of UEs. Group specific NAS level congestion control is performed at the 5GC only and it is transparent to UE. The AMF or SMF or both may apply NAS level congestion control for a UE associated to an Internal-Group Identifier (see clause 5.9.7).
NOTE: 5GC logic for Group specific NAS level congestion control is not described in this Release of the specification. |
fbecc7f0dcf9784c6066646052ab0c0e | 23.501 | 5.19.7.6 Control Plane data specific NAS level congestion control | Under overload conditions the AMF may restrict requests from UEs for data transmission via Control Plane CIoT 5GS Optimisation. A Control Plane data back-off timer may be returned by the AMF (e.g. in Registration Accept messages, Service Reject message or Service Accept message). While the Control Plane data back-off timer is running, the UE shall not initiate any data transfer via Control Plane CIoT 5GS Optimisation, i.e. the UE shall not send any Control Plane Service Request with uplink data as defined in TS 24.501 [47]. The AMF shall store the Control Plane data back-off timer per UE and shall not process any further requests (other than exception reporting and a response to paging) for Data Transport via a Control Plane Service Request from that UE while the Control Plane data back-off timer is still running.
NOTE 1: The Control Plane data back-off timer does not affect any other mobility management or session management procedure.
NOTE 2: The Control Plane data back-off timer does not apply to user plane data communication.
If the UE is allowed to send exception reporting, the UE may send an initial NAS Message for exception reporting even if Control Plane data back-off timer is running.
The UE may respond to paging with an initial NAS Message without uplink data even if the Control Plane data back-off timer is running.
If the AMF receives an initial NAS Message in response to a paging and the AMF has a Control Plane data back-off timer running for the UE and the AMF is not overloaded and AMF decides to accept the Control Plane Service Request, then the AMF shall respond with Service Accept without the Control Plane data back-off timer and stop the Control Plane data back-off timer. If the UE receives a Service Accept without the Control Plane data back-off timer from the AMF while the Control Plane data back-off timer is running, the UE shall stop the Control Plane data back-off timer. The Control Plane data back-off timer in the UE and the AMF is stopped at PLMN change.
If the AMF receives a Control Plane Service Request with uplink data and decides to send the UE a Control Plane data back-off timer, the AMF may decide to process the Control Plane Service Request with uplink data, i.e. decrypt and forward the data payload, or not based on the following:
- If the UE has indicated Release Assistance Information that no further Uplink and Downlink Data transmissions are expected, then the AMF may process (integrity check/decipher/forward) the received Control Plane data packet and send a Service Accept to the UE with Control Plane data back-off timer. The UE interprets this as successful transmission of the Control Plane data packet starts the Control Plane data back-off timer.
- For all other cases, the AMF may decide to not process the received Control Plane data packet and send a Service Reject to the UE with Control Plane data back-off timer. The UE interprets this indication as unsuccessful delivery of the control plane data packet and starts the Control Plane data back-off timer. The AMF may take into consideration whether the PDU Session is set to Control Plane only to make the decision whether to reject the packet and send Service Reject or move the PDU Session to user plane and process the data packet as described in next bullet.
- Alternatively, if UE has not provided Release Assistance Information and the PDU Session not set to Control Plane only and UE supports N3 data transfer, then the AMF may initiate establishment of N3 bearer according to the procedure defined in clause 4.2.3 of TS 23.502 [3]. In this case the AMF may also return a Control Plane data back-off timer within the Service Accept.
The AMF only includes the Control Plane data back-off timer if the UE has indicated support for Control Plane CIoT 5GS optimizations in the Registration Request. |
fbecc7f0dcf9784c6066646052ab0c0e | 23.501 | 5.20 External Exposure of Network Capability | The Network Exposure Function (NEF) supports external exposure of capabilities of network functions. External exposure can be categorized as Monitoring capability, Provisioning capability, Policy/Charging capability, Analytics reporting capability and Member UE selection capability. The Monitoring capability is for monitoring of specific event for UE in 5G System and making such monitoring events information available for external exposure via the NEF. The Provisioning capability is for allowing external party to provision of information which can be used for the UE in 5G System. The Policy/Charging capability is for handling access and mobility management, QoS and charging policies for the UE based on the request from external party. The Analytics reporting capability is for allowing an external party to fetch or subscribe/unsubscribe to analytics information generated by 5G System (this is further defined in TS 23.288 [86]). The Member UE selection capability is for allowing an external party to acquire one or more list(s) of candidate UE(s) (among the list of target member UE(s) provided by the AF) and additional information that is based on the assistance information generated by 5G System based on some defined filtering criteria, the details are explained in clause 4.15.13 in TS 23.502 [3].
Monitoring capability is comprised of means that allow the identification of the 5G network function suitable for configuring the specific monitoring events, detect the monitoring event and report the monitoring event to the authorised external party. Monitoring capability can be used for exposing UE's mobility management context such as UE location, reachability, roaming status and loss of connectivity. Monitoring capability can also be used for exposing QoS monitoring result. AMF stores URRP-AMF information in the MM context to determine the NFs that are authorised to receive direct notifications from the AMF. UDM stores URRP-AMF information locally to determine authorised monitoring requests when forwarding indirect notifications. The Monitoring capability also allows AF to subscribe to the group status changes for a group, either a 5G VN group as described in clause 5.29.2, as well as a group configured by OA&M. In this case the AF is notified if the group member list is updated or a group member is no longer subscribed to the group.
Provisioning capability allows an external party to provision the Expected UE Behaviour or the 5G-VN group information or DNN and S-NSSAI specific Group Parameters or ECS Address Configuration Information or service specific information or Static IP address assignment parameters to 5G NF via the NEF. The provisioning comprises of the authorisation of the provisioning external third party, receiving the provisioned external information via the NEF, storing the information and distributing that information among those NFs that use it. The externally provisioned data can be consumed by different NFs, depending on the data. In the case of provisioning the Expected UE Behaviour, the externally provisioned information which is defined as the Expected UE Behaviour parameters in clause 4.15.6.3 of TS 23.502 [3] or Network Control parameter in clause 4.15.6.3a of TS 23.502 [3] consists of information on expected UE movement, Expected UE Behaviour parameters or expected Network Configuration parameter. The provisioned Expected UE Behaviour parameters may be used for the setting of mobility management or session management parameters of the UE. In the case of provisioning the 5G-VN group information the externally provisioned information is defined as the 5G-VN group parameters in clause 4.15.6.7 of TS 23.502 [3] and it consists of some information on the 5G-VN group. In the case of the provisioning the DNN and S-NSSAI specific Group Parameters, the externally provisioned information is defined in clause 4.15.6.14 of TS 23.502 [3] and clause 5.20b. In the case of provisioning non-3GPP device Identifier information, the externally provisioned information is defined as the non-3GPP device Identifier information in clause 4.15.6.15 of TS 23.502 [3]. In the case of provisioning ECS address, the externally provisioned information is defined as the ECS Address Configuration Information in clause 4.15.6.3d of TS 23.502 [3]. The affected NFs are informed via the subscriber data update as specified in clause 4.15.6.2 of TS 23.502 [3]. The externally provisioned information which is defined as the Service Parameters in clause 4.15.6.7 of TS 23.502 [3] consists of service specific information used for supporting the specific service in 5G system. The provisioned Service Parameters may be delivered to the UEs. The affected NFs are informed of the data update.
Policy/Charging capability is comprised of means that allow the request for session and charging policy, enforce QoS policy, apply accounting functionality and requests to influence access and mobility management policies. It can be used for specific QoS/priority handling for the session of the UE and for setting applicable charging party or charging rate.
Analytics reporting capability is comprised of means that allow discovery of type of analytics that can be consumed by external party, the request for consumption of analytics information generated by NWDAF.
Member UE selection capability is comprised of means that allows filtering and providing one or more list(s) of candidate UE(s) (among the list of target member UE(s) provided by the AF) and additional information that can be consumed by external party, the request for consumption of UE list generated by external party.
An NEF may support CAPIF functions for external exposure as specified in clause 6.2.5.1.
An NEF may support exposure of NWDAF analytics as specified in TS 23.288 [86].
The NEF may support exposure of 5GS and/or UE availability and capabilities for time synchronization service as specified in clause 5.27.1.8.
An NEF may support exposure of event based notifications and reports for NSACF as specified in clause 5.15.11.
An NEF may support exposure of energy related information as specified in clause 5.51.2.
An AF may only be able to identify the target UE of an AF request for external exposure of 5GC capabilities (e.g. Data Provisioning or for Event Exposure for a specific UE) by providing the UE's address information. In this case the NEF first needs to retrieve the Permanent identifier of the UE before trying to fulfil the AF request. The NEF may determine the Permanent identifier of the UE, as described in clause 4.15.3.2.13 of TS 23.502 [3], based on:
- the address of the UE as provided by the AF; this may be an IP address or a MAC address;
- the corresponding DNN and/or S-NSSAI information: this may have been provided by the AF or determined by the NEF based on the requesting AF; this is needed if the UE address is an IP address.
The NEF may provide a UE Identifier in the GPSI format of MSISDN to an authenticated and authorized AF:
- that fulfils the conditions described in clause 4.15.10A of TS 23.502 [3]; and
- that has explicitly requested a translation from the UE address to a unique UE identifier (via Nnef_UEId service) when the UE MSISDN exposure is allowed and authorized by the operator; or
- the NEF may provide an AF specific UE Identifier to the AF:
- that has explicitly requested a translation from the address of the UE to a unique UE identifier (via Nnef_UEId service); or
- that has implicitly requested a translation from the address of the UE to an AF specific UE Identifier by requesting external exposure about an individual UE identified by its address.
The AF may have its own means to maintain the AF specific UE Identifier through, e.g. an AF session. After the retrieval of an AF specific UE Identifier the AF shall not keep maintaining a mapping between this identifier and the UE IP address as this mapping may change.
The AF specific UE Identifier shall not correspond to a MSISDN; it is represented as a GPSI in the form of an External Identifier. When used as an AF specific UE identifier, the External Identifier provided by the 5GCN shall be different for different AF.
NOTE 1: This is to protect user privacy.
NOTE 2: The AF specific UE identifier is ensured to be unique across different AFs as defined in TS 23.003 [19] by configuration. Such configuration is assumed to be coordinated between the different involved entities (e.g. NEF(s) and UDM/UDR).
NOTE 3: Based on policies, the NEF can be configured to enforce restriction on the usage of AF specific UE identifier (e.g. rejection of a service request from AF not authorized to use the UE identifier).
The NEF may also provide MSISDN verification (via Nnef_UEId service) to an authenticated and authorized AF that fulfils the conditions described in clause 4.15.10B of TS 23.502 [3].
5.20a Data Collection from an AF
An NF that needs to collect data from an AF may subscribe/unsubscribe to notifications regarding data collected from an AF, either directly from the AF or via NEF.
The data collected from an AF is used as input for analytics by the NWDAF.
The details for the data collected from an AF as well as interactions between NEF, AF and NWDAF are described in TS 23.288 [86].
5.20b Support exposure of DNN and S-NSSAI specific Group Parameters
5.20b.1 Group attribute provisioning
A group may be a 5G VN group managed as defined in clause 5.29.2, as well as a group configured by OA&M.
An AF may provision DNN and S-NSSAI specific attributes for a group of UEs:
- LADN Service area, which consists of Tracking Area identities or geographical information, it is applicable to each UE member within the group and for a specific DNN and S-NSSAI.
- Default QoS, the QoS refers to 5QI, ARP and 5QI Priority Level as defined in clause 5.7.2.7 and it is applicable to each UE member within the group and for a specific DNN and S-NSSAI.
5.20b.2 Support LADN service area for a group
The procedure as defined in clause 4.15.6.2 of TS 23.502 [3] is applicable for provisioning of LADN service area for a group with the following clarifications and enhancements:
- The AF request additionally contains the LADN service area as part of DNN and S-NSSAI specific Group Parameters and the LADN service area is stored in UDR as subscription data and delivered to AMF. If the AMF receives the LADN service area for a group, the AMF configures the DNN of the group as LADN DNN.
- If the AF provides the LADN service area in the form of geographical information, the NEF maps the geographical information to a list of TAs before sending the service area to the UDM.
LADN per DNN and S-NSSAI as defined in clause 5.6.5a is applicable for enforcement of LADN service area.
5.20b.3 Support QoS for a group
The procedure as defined in clause 4.15.6.2 of TS 23.502 [3] is applicable for provisioning of default QoS for a DNN and S-NSSAI for a group of UEs with the following clarifications and enhancements:
- The AF request contains the Default QoS for the group and the UDM stores such QoS in the UDR and uses such QoS to set 5GS Subscribed QoS profile in Session Management Subscription data for each UE within the group.
NOTE: When a UE belongs to multiple groups simultaneously and AF(s) provision different Default QoS for the same DNN and S-NSSAI but different groups, UDM selects a QoS profile among the groups the UE belongs to for a DNN and S-NSSAI to set the 5GS subscribed QoS profile. How the UDM selects a QoS profile is based on implementation and configuration. The UDM can e.g. select a QoS profile with a higher 5QI Priority Level or higher ARP priority level.
Mechanisms as defined in clause 5.7.2.7 are used to enforce the 5GS Subscribed QoS profile for a DNN and S-NSSAI for each UE within a group.
5.20b.4 Void
5.20b.5 Void
5.20c Provisioning of traffic characteristics and monitoring of performance characteristics for a group
NEF provisioning capability as defined in clause 5.20 allows an AF to perform provisioning of traffic characteristics and monitoring of performance characteristics for a UE or group of UEs as specified in clause 4.15.6.14 of TS 23.502 [3] and clause 6.1.3.28 of TS 23.503 [45].
NOTE : The AF may use application layer functionalities to handle requests for UE-to-UE traffic as defined by SA WG6.
The NEF determines whether or not to invoke the TSCTSF in the same way as for AF session with required QoS procedure, as described in step 2 of clause 4.15.6.6 in TS 23.502 [3]. In the case that the TSCTSF is used, the TSCTSF receives the AF requested QoS information from the NEF and handles it as described in clause 4.15.6.14 of TS 23.502 [3]. In the case that TSCTSF is not used, the AF request is handled as described in clause 4.15.6.14 of TS 23.502 [3] and clause 6.1.3.28 of TS 23.503 [45].
When the TSCTSF receives the AF requested QoS information from NEF or the PCF(s) receive the AF requested QoS information from UDR, the TSCTSF or PCF (s) manage the AF requested QoS information for each UE group member within the group as follows:
- TSCTSF translates the GPSI or External/Internal Group ID into a list of SUPIs by invoking Nudm_SDM_Get service.
- Determine which of these UE group members have active PDU Sessions matching the DNN and S-NSSAI and determine the relevant UE address.
- Manage the request status (activated, de-activated, failed) for each UE group member within the group:
- Apply the AF requested QoS information if the UE group member is registered or has active PDU Session matching the DNN and S-NSSAI. Set the status to activated if the QoS resources are assigned for the UE group member, otherwise, failed.
- Set the status to de-activated if the UE group member is not registered or has no active PDU Session matching the DNN and S-NSSAI.
- Delete the request status for the UE group member if the UE group member is removed from the group and further revokes AF request QoS information if the request status is activated.
- Check whether to apply the AF requested QoS information and update the request status for the UE group member if the UE group member is newly added to the group.
- When the AF requested QoS information contains temporal invalidity condition as described in clause 6.1.3.28 of TS 23.503 [45]:
- Revoke AF requested QoS information for each UE group member which is marked as active, so to e.g. to remove or modify PCC rules as defined in clause 4.16.5.2 of TS 23.502 [3] if the start-time is reached.
- Apply AF requested QoS information for each UE group member that has active PDU Sessions matching the DNN and S-NSSAI, e.g. to add or modify PCC rules as defined in clause 4.16.5.2 of TS 23.502 [3] if the end-time is reached.
5.20d User Plane Direct 5GS Information Exposure
5.20d.1 General
In order to expose network information, the user plane direct 5GS information exposure function may be applied. The user plane direct 5GS information exposure function allows the UPF to report the network information directly to consumer based on the instructions provided by SMF.
NOTE: In the scenario of Edge Computing as described in TS 23.548 [130], the consumer can be the L-NEF or local AF when the local AF is trusted.
When the exposed network information is provided by the UPF, the PSA UPF may be instructed to report network information via Nupf_EventExposure service (e.g. directly to an AF, i.e. bypassing the SMF and the PCF); or the UPF may be instructed to report the information to the consumer via SMF/PCF/NEF, as described in clause 5.8.2.18.
When the exposed network information is provided by the NG-RAN, the NG-RAN may be instructed by the SMF to report the information via the GTP-U tunnel(s) between the NG RAN and PSA UPF, as defined in clause 5.45.
The User Plane Direct 5GS Information Exposure may be used for exposing the following information:
- QoS Monitoring information (see clause 5.45).
- TSC Management Information (see clause 5.28.3). |
fbecc7f0dcf9784c6066646052ab0c0e | 23.501 | 5.21 Architectural support for virtualized deployments | |
fbecc7f0dcf9784c6066646052ab0c0e | 23.501 | 5.21.0 General | 5GC supports different virtualized deployment scenarios, including but not limited to the options below:
- A Network Function instance can be deployed as distributed, redundant, stateless and scalable NF instance that provides the services from several locations and several execution instances in each location.
- This type of deployments would typically not require support for addition or removal of NF instances for redundancy and scalability. In the case of an AMF this deployment option may use enablers like, addition of TNLA, removal of TNLA, TNLA release and rebinding of NGAP UE association to a new TNLA to the same AMF.
- A Network Function instance can also be deployed such that several network function instances are present within a NF set provide distributed, redundant, stateless and scalability together as a set of NF instances.
- This type of deployments may support for addition or removal of NF instances for redundancy and scalability. In the case of an AMF this deployment option may use enablers like, addition of AMFs and TNLAs, removal of AMFs and TNLAs, TNLA release and rebinding of NGAP UE associations to a new TNLA to different AMFs in the same AMF set.
- The SEPP, although not a Network Function instance, can also be deployed distributed, redundant, stateless and scalable.
- The SCP, although not a Network Function instance, can also be deployed distributed, redundant and scalable.
Also, deployments taking advantage of only some or any combination of concepts from each of the above options is possible. |
fbecc7f0dcf9784c6066646052ab0c0e | 23.501 | 5.21.1 Architectural support for N2 | |
fbecc7f0dcf9784c6066646052ab0c0e | 23.501 | 5.21.1.1 TNL associations | 5G-AN node shall have the capability to support multiple TNL associations per AMF, i.e. AMF name.
An AMF shall provide the 5G-AN node with the weight factors for each TNL association of the AMF.
The AMF shall be able to request the 5G-AN node to add or remove TNL associations to the AMF.
The AMF shall be able to indicate to the 5G-AN node the set of TNL associations used for UE-associated signalling and the set of TNL associations used for non-UE associated signalling.
NOTE: The TNL association(s) indicated for UE-associated and non-UE associated signalling can either be overlap or be different. |
fbecc7f0dcf9784c6066646052ab0c0e | 23.501 | 5.21.1.2 NGAP UE-TNLA-binding | While a UE is in CM-Connected state the 5G-AN node shall maintain the same NGAP UE-TNLA-binding (i.e. use the same TNL association and same NGAP association for the UE) unless explicitly changed or released by the AMF.
An AMF shall be able to update the NGAP UE-TNLA-binding (i.e. change the TNL association for the UE) in CM-CONNECTED state at any time. The NGAP UE-TNLA-binding can also be updated when a UE-specific NGAP message initiated by AMF is received via a new TNL association.
An AMF shall be able to update the NGAP UE-TNLA-binding (i.e. change the TNL association for the UE) in response to an N2 message received from the 5G-AN by triangular redirection (e.g. by responding to the 5G-AN node using a different TNL association).
An AMF shall be able to command the 5G-AN node to release the NGAP UE-TNLA-binding for a UE in CM-CONNECTED state while maintaining N3 (user-plane connectivity) for the UE at any time. |
fbecc7f0dcf9784c6066646052ab0c0e | 23.501 | 5.21.1.3 N2 TNL association selection | The 5G-AN node shall consider the following factors for selecting a TNL association for the AMF for the initial N2 message e.g. N2 INITIAL UE MESSAGE:
- Availability of candidate TNL associations.
- Weight factors of candidate TNL associations.
The AMF may use any TNL association intended for non-UE associated signalling for initiation of the N2 Paging procedure. |
fbecc7f0dcf9784c6066646052ab0c0e | 23.501 | 5.21.2 AMF Management | |
fbecc7f0dcf9784c6066646052ab0c0e | 23.501 | 5.21.2.1 AMF Addition/Update | The 5G System should support establishment of association between AMF and 5G-AN node.
A new AMF can be added to an AMF set and association between AMF and GUAMI can be created and/or updated as follows:
- AMF shall be able to dynamically update the NRF with the new or updated GUAMI(s) to provide mapping between GUAMI(s) and AMF information. Association between GUAMI(s) and AMF is published to NRF. In addition, to deal with planned maintenance and failure, an AMF may optionally provide backup AMF information, i.e. it act as a backup AMF if the indicated GUAMI associated AMF is unavailable. It is assumed that the backup AMF and the original AMF are in the same AMF set as they have access to the same UE context. Based on that information one GUAMI is associated with an AMF, optionally with a backup AMF used for planned removal and/or another (same or different) backup AMF used for failure.
- Upon successful update, the NRF considers the new and/or updated GUAMI(s) for providing AMF discovery results to the requester. Requester can be other CP network functions.
- The new AMF provides its GUAMI to 5G-AN and 5G-AN store this association. If the association between the same GUAMI and another AMF exists in the 5G-AN (e.g. due to AMF planned removal), the previously stored AMF is replaced by the new AMF for the corresponding GUAMI association.
Information about new AMF should be published and available in the DNS system. It should allow 5G-AN to discover AMF and setup associations with the AMF required. N2 setup procedure should allow the possibility of AMFs within the AMF Set to advertise the same AMF Pointer and/or distinct AMF Pointer value(s) to the 5G-AN node.
To support the legacy EPC core network entity (i.e. MME) to discover and communicate with the AMF, the information about the AMF should be published and available in the DNS system. Furthermore, GUMMEI and GUAMI encoding space should be partitioned to avoid overlapping values in order to enable MME discover an AMF without ambiguity. |
fbecc7f0dcf9784c6066646052ab0c0e | 23.501 | 5.21.2.2 AMF planned removal procedure | |
fbecc7f0dcf9784c6066646052ab0c0e | 23.501 | 5.21.2.2.1 AMF planned removal procedure with UDSF deployed | An AMF can be taken graciously out of service as follows:
- If an UDSF is deployed in the network, then the AMF stores the context for registered UE(s) in the UDSF. The UE context includes the AMF UE NGAP ID that is unique per AMF set. In order for the AMF planned removal procedure to work graciously, 5G-S-TMSI shall be unique per AMF Set. If there are ongoing transactions (e.g. N1 procedure) for certain UE(s), AMF stores the UE context(s) in the UDSF upon completion of an ongoing transaction.
- The AMF deregister itself from NRF indicating due to AMF planned removal.
NOTE 1: It is assumed that the UE contexts from the old AMF include all event subscriptions with peer CP NFs.
NOTE 2: Before removal of AMF the overload control mechanism can be used to reduce the amount of ongoing transaction.
An AMF identified by GUAMI(s) shall be able to notify the 5G-AN that it will be unavailable for processing transactions by including GUAMI(s) configured on this AMF. Upon receipt of the indication that an AMF(identified by GUAMI(s)) is unavailable, 5G-AN shall take the following action:
- 5G-AN should mark this AMF as unavailable and not consider the AMF for selection for subsequent N2 transactions until 5G-AN learns that it is available (e.g. as part of discovery results or by configuration).
- During NGAP Setup procedure, the AMF may include an additional indicator that the AMF will rebind or release the NGAP UE-TNLA-binding on a per UE-basis for UE(s) in CM-CONNECTED state. If that indicator is included and the 5G-AN supports timer mechanism, the 5G-AN starts a timer to control the release of NGAP UE-TNLA-binding. For the duration of the timer or until the AMF releases or re-binds the NGAP UE-TNLA-binding the AN does not select a new AMF for subsequent UE transactions. Upon timer expiry, the 5G-AN releases the NGAP UE UE-TNLA-binding(s) with the corresponding AMF for the respective UE(s), for subsequent N2 message, the 5G-AN should select a different AMF from the same AMF set when the subsequent N2 message needs to be sent.
NOTE 3: For UE(s) in CM-CONNECTED state, after indicating that the AMF is unavailable for processing UE transactions and including an indicator that the AMF releases the NGAP UE-TNLA-binding(s) on a per UE-basis, the AMF can either trigger a re-binding of the NGAP UE associations to an available TNLA on a different AMF in the same AMF set or use the NGAP UE-TNLA-binding per UE release procedure defined in TS 23.502 [3] to release the NGAP UE-TNLA-binding on a per UE-basis while requesting the AN to maintain N3 (user plane connectivity) and UE context information.
NOTE 4: The support and the use of timer mechanism in 5G-AN is up to implementation.
- If the instruction does not include the indicator, for UE(s) in CM-CONNECTED state, 5G-AN considers this as a request to release the NGAP UE-TNLA-binding with the corresponding AMF for the respective UE(s) while maintaining N3 (user plane connectivity) and UE context information. For subsequent N2 message, the 5G-AN should select a different AMF from the same AMF set when the subsequent N2 message needs to be sent.
- For UE(s) in CM-IDLE state, when it subsequently returns from CM-IDLE state and the 5G-AN receives an initial NAS message with a 5G S-TMSI or GUAMI pointing to an AMF that is marked unavailable, the 5G-AN should select a different AMF from the same AMF set and forward the initial NAS message. If the 5G-AN can't select an AMF from the same AMF set, the 5G-AN selects another new AMF as described in clause 6.3.5.
An AMF identified by GUAMI(s) shall be able to instruct other peer CP NFs, subscribed to receive such a notification, that it will be unavailable for processing transactions by including GUAMI(s) configured on this AMF. If the CP NFs register with NRF for AMF unavailable notification, then the NRF shall be able to notify the subscribed NFs to receive such a notification that AMF identified by GUAMI(s) will be unavailable for processing transactions. Upon receipt of the notification that an AMF (GUAMI(s)) is unavailable, the other CP NFs shall take the following actions:
- CP NF should mark this AMF (identified by GUAMI(s)) as unavailable and not consider the AMF for selection for subsequent MT transactions until the CP NF learns that it is available (e.g. as part of NF discovery results or via NF status notification from NRF).
- Mark this AMF as unavailable while not changing the status of UE(s) associated to this AMF (UE(s) previously served by the corresponding AMF still remain registered in the network) and AMF Set information.
- For the UE(s) that were associated to the corresponding AMF, when the peer CP NF needs to initiate a transaction towards the AMF that is marked unavailable, CP NF should select another AMF from the same AMF set (as in clause 6.3.5) and forward the transaction together with the old GUAMI. The new AMF retrieves UE context from the UDSF. If CP NF needs to send a notification to new AMF which is associated with a subscription from the old AMF, the CP NF shall exchange the old AMF information embedded in the Notification Address with the new AMF information and use that Notification Address for subsequent communication.
NOTE 5: If the CP NF does not subscribe to receive AMF unavailable notification (either directly from the AMF or via NRF), the CP NF may attempt forwarding the transaction towards the old AMF and detect that the AMF is unavailable after certain number of attempts. When it detects unavailable, it marks the AMF and its associated GUAMI(s) as unavailable. CP NF should select another AMF from the same AMF set (as in clause 6.3.5) and forward the transaction together with the old GUAMI. The new AMF retrieves UE context from the UDSF and process the transaction.
Following actions should be performed by the newly selected AMF:
- When there is a transaction with the UE the newly selected AMF retrieves the UE context from the UDSF based on SUPI, 5G-GUTI or AMF UE NGAP ID and processes the UE message accordingly and updates the 5G-GUTI towards the UE, if necessary. For UE(s) in CM-CONNECTED state, it may also update the NGAP UE association with a new AMF UE NGAP ID towards the 5G-AN and replace the GUAMI in the UE context stored at the 5G-AN with the new GUAMI associated with the newly selected AMF if the 5G-GUTI has been updated. The AMF also informs the NG-RAN of the new UE Identity Index Value (derived from the new 5G-GUTI).
- When there is a transaction with the UE, the new selected AMF updates the peer NFs (that subscribed to receive AMF unavailability notification from old AMF), with the new selected AMF information.
- If the new AMF is aware of a different AMF serving the UE (by implementation specific means) it forwards the uplink N2 signalling of the UE to that AMF directly if necessary, the 5G-AN shall be able to receive the message from a different AMF, or it rejects the transaction from the peer CP NFs with a cause to indicate that new AMF has been selected, the peer CP NFs resend the transaction to the new AMF.
NOTE 6: This bullet above addresses situations where 5G-AN node selects an AMF and CP NFs select another AMF for the UE concurrently. It also addresses the situation where CP NFs select an AMF for the UE concurrently
- If the UE is in CM-IDLE state and the new AMF does not have access to the UE context, the new AMF selects one available AMF from the old AMF set as described in clause 6.3.5. The selected AMF retrieves the UE context from the UDSF and provides the UE context to the new AMF. If the new AMF doesn't receive the UE context then the AMF may force the UE to perform Initial Registration. |
fbecc7f0dcf9784c6066646052ab0c0e | 23.501 | 5.21.2.2.2 AMF planned removal procedure without UDSF | An AMF can be taken graciously out of service as follows:
- The AMF can forward registered UE contexts, UE contexts grouped by the same GUAMI value, to target AMF(s) within the same AMF set, including the source AMF name used for redirecting UE's MT transaction. The UE context includes the per AMF Set unique AMF UE NGAP ID. In order for the AMF planned removal procedure to work graciously, 5G-S-TMSI shall be unique per AMF set. If there are ongoing transactions (e.g. N1 procedure) for certain UE(s), AMF forwards the UE context(s) to the target AMF upon completion of an ongoing transaction.
- The AMF deregister itself from NRF indicating due to AMF planned removal.
NOTE 1: It is assumed that the UE contexts from the old AMF include all event subscriptions with peer CP NFs.
NOTE 2: Before removal of AMF the overload control mechanism can be used to reduce the amount of ongoing transaction.
An AMF shall be able to instruct the 5G-AN that it will be unavailable for processing transactions by including GUAMI(s) configured on this AMF and its corresponding target AMF(s). The target AMF shall be able to update the 5G-AN that the UE(s) served by the old GUAMI(s) are now served by target AMF. The target AMF provides the old GUAMI value that the 5G-AN can use to locate UE contexts served by the old AMF. Upon receipt of the indication that an old AMF is unavailable, 5G-AN shall take the following action:
- 5G-AN should mark this AMF as unavailable and not consider the AMF for selection for subsequent N2 transactions until 5G-AN learns that it is available (e.g. as part of discovery results or by configuration). The associated GUAMIs are marked as unavailable.
- During NGAP Setup, the AMF may include an additional indicator that the AMF will rebind or release the NGAP UE-TNLA-binding on per UE-basis. If that indicator is included and the 5G-AN supports timer mechanism, the 5G-AN starts a timer to control the release of NGAP UE-TNLA-binding(s). For the duration of the timer or until the AMF releases or re-binds the NGAP UE-TNLA-binding, the AN does not select a new AMF for subsequent transactions. Upon timer expiry, the 5G-AN releases the NGAP UE-TNLA-binding(s) with the corresponding AMF for the respective UE(s), for subsequent N2 message, the 5G-AN uses GUAMI which points to the target AMF that replaced the old unavailable AMF, to forward the N2 message to the corresponding target AMF(s).
NOTE 3: For UE(s) in CM-CONNECTED state, after indicating that the AMF is unavailable for processing UE transactions and including an indicator that the AMF releases the NGAP UE-TNLA-binding on a per UE-basis, the AMF can either trigger a re-binding of the NGAP UE associations to an available TNLA on a different AMF within the same AMF set or use the NGAP UE-TNLA-binding per UE release procedure defined in TS 23.502 [3] to release the NGAP UE-TNLA-binding on a per UE-basis while requesting the AN to maintain N3 (user plane connectivity) and UE context information.
NOTE 4: The support and the use of timer mechanism in 5G-AN is up to implementation.
If the instruction does not include the indicator, for UE(s) in CM-CONNECTED state, 5G-AN considers this as a request to release the NGAP UE UE-TNLA-binding(s) with the corresponding AMF for the respective UE(s) while maintaining N3 (user plane connectivity) and UE context information. For subsequent N2 message, the 5G-AN uses GUAMI based resolution which points to the target AMF that replaced the old unavailable AMF, to forward the N2 message to the corresponding target AMF(s).
- For UE(s) in CM-IDLE state, when it subsequently returns from CM-IDLE state and the 5G-AN receives an initial NAS message with a 5G S-TMSI or GUAMI, based resolution the 5G-AN uses 5G S-TMSI or GUAMI which points to the target AMF that has replaced the old unavailable AMF and, the 5G-AN forwards N2 message.
An AMF shall be able to instruct other peer CP NFs, subscribed to receive such a notification, that it will be unavailable for processing transactions by including GUAMI(s) configured on this AMF and its corresponding target AMF(s). The target AMF shall update the CP NF that the old GUAMI(s) is now served by target AMF. The old AMF provides the old GUAMI value to target AMF and the target AMF can use to locate UE contexts served by the old AMF. If the CP NFs register with NRF for AMF unavailable notification, then the NRF shall be able to notify the subscribed NFs to receive such a notification (along with the corresponding target AMF(s)) that AMF identified by GUAMI(s) will be unavailable for processing transactions. Upon receipt of the notification that an AMF is unavailable, the other CP NFs shall take the following action:
- Mark this AMF and its associated GUAMI(s) as unavailable while not changing the status of UE(s) associated to this AMF (UE(s) previously served by the corresponding AMF still remain registered in the network) and AMF Set information.
- For the UE(s) that were associated to the corresponding AMF, when the peer CP NF needs to initiate a transaction towards the AMF that is marked unavailable and the old unavailable AMF was replaced by the target AMF, CP NF should forward the transaction together with the old GUAMI to the target AMF(s). If CP NF needs to send a notification to new AMF which is associated with a subscription from the old AMF, the CP NF shall exchange the old AMF information embedded in the Notification Address with the new AMF information and use that Notification Address for subsequent communication.
NOTE 5: If the CP NF does not subscribe to receive AMF unavailable notification (either directly with the AMF or via NRF), the CP NF may attempt forwarding the transaction towards the old AMF and detect that the AMF is unavailable after certain number of attempts. When it detects unavailable, it marks the AMF and its associated GUAMI(s) as unavailable.
The following actions should be performed by the target AMF:
- To allow AMF process ongoing transactions for some UE(s) even after it notifies unavailable status to the target AMF, the target AMF keeps the association of the old GUAMI(s) and the old AMF for a configured time. During that configured period, if target AMF receives the transaction from the peer CP NFs and cannot locate UE context, it rejects the transaction with old AMF name based on that association and the indicated AMF is only used for the ongoing transaction. The peer CP NFs resend the transaction to the indicated AMF only for the ongoing transaction. For subsequent transactions, peer CP NFs should use the target AMF. When the timer is expired, the target AMF deletes that association information.
- When there is a transaction with the UE the target AMF uses SUPI, 5G-GUTI or AMF UE NGAP ID to locate UE contexts and processes the UE transactions accordingly and updates the 5G-GUTI towards the UE, if necessary. For UE(s) in CM-CONNECTED state, it may also update the NGAP UE association with a new AMF UE NGAP ID towards the 5G-AN and replace the GUAMI in the UE context stored at the 5G-AN with the new GUAMI associated with the newly selected AMF if the 5G-GUTI has been updated. The AMF also informs the NG-RAN of the new UE Identity Index Value (derived from the new 5G-GUTI).
- Target AMF shall not use old GUAMI to allocate 5G-GUTI for UE(s) that are being served by Target AMF. |
fbecc7f0dcf9784c6066646052ab0c0e | 23.501 | 5.21.2.3 Procedure for AMF Auto-recovery | In order to try and handle AMF failure in a graceful manner (i.e. without impacting the UE), AMF can either back up the UE contexts in UDSF, or per GUAMI granularity in other AMFs (serving as backup AMF for the indicated GUAMI).
NOTE 1: Frequency of backup is left to implementation.
For deployments without UDSF, for each GUAMI the backup AMF information (in association to the GUAMI) is configured in the AMF. The AMF sends this information to 5G-AN and other CP NFs during the N2 setup procedure or the first (per NF) interaction with other CP NFs.
In the case that an AMF fails and the 5G-AN/peer CP NFs detect that the AMF has failed, or the 5G-AN/peer CP NFs receives notification from another AMF in the same AMF set that this AMF has failed, following actions are taken:
- The OAM deregister the AMF from NRF indicating due to AMF failure.
- 5G-AN marks this AMF as failed and not consider the AMF for selection until explicitly notified.
- For UE(s) in CM-CONNECTED state, 5G-AN considers failure detection or failure notification as a trigger to release the NGAP UE-TNLA-binding(s) with the corresponding AMF for the respective UE(s) while maintaining N3 (user plane connectivity) and other UE context information. For subsequent N2 message, if the backup AMF information of the corresponding failed AMF is not available the 5G-AN should select a different AMF (as in clause 6.3.5) from the same AMF set when the subsequent N2 message needs to be sent for the UE(s). If no other AMF from the AMF set is available, then it can select an AMF (implementation dependent) from the same AMF Region as in clause 6.3.5. If backup AMF information of the corresponding failed AMF is available, the 5G-AN forwards the N2 message to the backup AMF.
NOTE 2: One AMF in the AMF set may be configured to send this failure notification message.
- For UE(s) in CM-IDLE state, when it subsequently returns from CM-IDLE state and the 5G-AN receives an initial NAS message with a S-TMSI or GUAMI pointing to an AMF that is marked failed, if the backup AMF information of the corresponding failed AMF is not available the 5G-AN should select a different AMF from the same AMF set and forward the initial NAS message. If no other AMF from the AMF set is available, then it can select an AMF (implementation dependent) from the same AMF Region as in clause 6.3.5. If backup AMF information of the corresponding failed AMF is available, the 5G-AN forwards the N2 message to the backup AMF.
- Peer CP NFs consider this AMF as unavailable while retaining the UE context.
- For the UE(s) that were associated to the corresponding AMF, when the peer CP NF needs to initiate a transaction towards the AMF, if backup AMF information of the corresponding failed AMF is not available, CP NF should select another AMF from the same AMF set and forward the transaction together with the old GUAMI. If neither the backup AMF nor any other AMF from the AMF set is available, then CP NF can select an AMF from the same AMF Region as in clause 6.3.5. If backup AMF information of the corresponding failed AMF is available, the CP NF forwards transaction to the backup AMF. If CP NF needs to send a notification to new AMF which is associated with a subscription from the old AMF, the CP NF shall exchange the old AMF information embedded in the Notification Address with the new AMF information and use that Notification Address for subsequent communication.
- When the 5G-AN or CP NFs need to select a different AMF from the same AMF set,
- For deployments with UDSF, any AMF from the same AMF set can be selected.
- For deployments without UDSF, the backup AMF is determined based on the GUAMI of the failed AMF.
Following actions should be taken by the newly selected AMF:
- For deployments with UDSF, when there is a transaction with the UE the newly selected AMF retrieves the UE context from the UDSF using SUPI, 5G-GUTI or AMF UE NGAP ID and it processes the UE message accordingly and updates the 5G-GUTI towards the UE, if necessary.
- For deployments without UDSF, backup AMF (the newly selected AMF), based on the failure detection of the old AMF, instructs peer CP NFs and 5G-AN that the UE contexts corresponding to the GUAMI of the failed AMF is now served by this newly selected AMF. The backup AMF shall not use old GUAMI to allocate 5G-GUTI for UE(s) that are being served by Target AMF. The backup AMF uses the GUAMI to locate the respective UE Context(s).
- When there is a transaction with the UE, the new AMF updates the peer NFs (that subscribed to receive AMF unavailability notification from old AMF) with the new AMF information.
- If the new AMF is aware of a different AMF serving the UE (by implementation specific means) it redirects the uplink N2 signalling to that AMF, or reject the transaction from the peer CP NFs with a cause to indicate that new AMF has been selected. The peer CP NFs may wait for the update from the new AMF and resend the transaction to the new AMF.
NOTE 3: This bullet above addresses situations where 5G-AN node selects an AMF and other CP NFs select an AMF for the UE concurrently. It also addresses the situation where CP NFs select an AMF for the UE concurrently.
NOTE 4: It is assumed that the UE contexts from the old AMF include all event subscriptions with peer CP NFs.
- If the UE is in CM-IDLE state and the new AMF does not have access to the UE context, the new AMF selects one available AMF from the old AMF set as described in clause 6.3.5. The selected AMF retrieves the UE context from the UDSF and provides the UE context to the new AMF. If the new AMF doesn't receive the UE context then the AMF may force the UE to perform Initial Registration.
- If the UE is in CM-CONNECTED state, the new AMF may also update the NGAP UE association with a new AMF UE NGAP ID towards the 5G-AN and replace the GUAMI in the UE context stored at the 5G-AN with the new GUAMI associated with the newly selected AMF if the 5G-GUTI has been updated.
NOTE 5: The above N2 TNL association selection and AMF management is applied to the selected PLMN. |
fbecc7f0dcf9784c6066646052ab0c0e | 23.501 | 5.21.3 Network Reliability support with Sets | |
fbecc7f0dcf9784c6066646052ab0c0e | 23.501 | 5.21.3.1 General | A Network Function instance can be deployed such that several network function instances are present within an NF Set to provide distribution, redundancy and scalability together as a Set of NF instances. The same is also supported for NF Services. This can be achieved when the equivalent NFs and NF Services share the same context data or by Network Function/NF Service Context Transfer procedures as specified in clause 4.26 of TS 23.502 [3].
NOTE: A NF can be replaced by an alternative NF within the same NF Set in the case of scenarios such as failure, load balancing, load re-balancing.
Such a network reliability design shall work in both communication modes, i.e. Direct Communication and Indirect Communication. In the Direct Communication mode, the NF Service consumer is involved in the reliability related procedures. In Indirect Communication mode, the SCP is involved in the reliability related procedures. |
fbecc7f0dcf9784c6066646052ab0c0e | 23.501 | 5.21.3.2 NF Set and NF Service Set | Equivalent Control Plane NFs may be grouped into NF Sets, e.g. several SMF instances are grouped into an SMF Set. NFs within a NF Set are interchangeable because they share the same context data and may be deployed in different locations, e.g. different data centres.
In the case of SMF, multiple instances of SMFs within an SMF Set need to be connected to the same UPF:
- If the N4 association is established between a SMF instance and an UPF, each N4 association is only managed by the related SMF instance.
- If only one N4 association is established between a SMF Set and an UPF, any SMF in the SMF Set should be able to manage the N4 association with the UPF.
Furthermore, for a given UE and PDU Session any SMF in the SMF Set should be able to control the N4 session with the UPF (however, at any given time, only one SMF in the SMF Set will control the UPF for a given UE's PDU Session).
A Control Plane NF is composed of one or multiple NF Services. Within a NF a NF service may have multiple instances. These multiple NF Service instances can be grouped into NF Service Set if they are interchangeable with each other because they share the same context data.
NOTE: The actual mapping of instances to a given Set is up to deployment. |
fbecc7f0dcf9784c6066646052ab0c0e | 23.501 | 5.21.3.3 Reliability of NF instances within the same NF Set | The NF producer instance is the NF instance which host the NF Service Producer. When the NF producer instance is not available, another NF producer instance within the same NF Set is selected.
For Direct Communication mode, the NF Service consumer may subscribe to status change notifications of NF instance from the NRF. If the NF Service consumer is notified by the NRF or detects by itself (e.g. request is not responded) that the NF producer instance is not available anymore, another available NF producer instance within the same NF Set is selected by the NF Service consumer.
For Indirect Communication mode, the SCP or NF Service consumer may subscribe to status change notifications of NF instance from the NRF and select another NF producer instance within the same NF Set if the original NF producer instance serving the UE is not available anymore.
For both Direct Communication mode, the NF, or Indirect Communication mode, the NF or SCP may select another NF producer instance within the same NF Set based on notification of subscription of signalling storm analytics from NWDAF.
NOTE: It is up to the implementation on how the SCP knows a NF producer instance is not available anymore. |
fbecc7f0dcf9784c6066646052ab0c0e | 23.501 | 5.21.3.4 Reliability of NF Services | When multiple NF Service instances within a NF Service Set are exposed to the NF Service consumer or SCP and the failure of NF Service instance is detected or notified by the NRF, i.e. it is not available anymore, the NF Service consumer or SCP selects another NF Service instance of the same NF Service Set within the NF instance, if available. Otherwise the NF Service consumer or SCP selects a different NF instance within the same NF Set. The NF service consumer or SCP may also take the above action based on notification of subscription of signalling storm analytics from NWDAF.
NOTE: The NF Producer instance can change the NF Service instance in the response to the service request.
When multiple NF Service instances within a NF Service Set are exposed to the NF Service consumer or SCP as a single NF Service, the reliability, i.e. the selection of an alternative NF Service instance is handled within the NF instance. |
fbecc7f0dcf9784c6066646052ab0c0e | 23.501 | 5.21.4 Network Function/NF Service Context Transfer | |
fbecc7f0dcf9784c6066646052ab0c0e | 23.501 | 5.21.4.1 General | Network Function/NF Service Context Transfer Procedures allow transfer of Service Context of a NF/NF Service from a Source NF/NF Service Instance to the Target NF/NF Service Instance e.g. before the Source NF/NF Service can gracefully close its NF/NF Service. Service Context Transfer procedures are supported as specified in clause 4.26 of TS 23.502 [3].
Source NF / OA&M system determines when Source NF needs to transfer UE contexts to an NF in another NF set. Source NF should initiate this only for UE(s) that are not active in order to limit and avoid impacting services offered to corresponding UE(s). |
fbecc7f0dcf9784c6066646052ab0c0e | 23.501 | 5.22 System Enablers for priority mechanism | |
fbecc7f0dcf9784c6066646052ab0c0e | 23.501 | 5.22.1 General | The 5GS and the 5G QoS model allow classification and differentiation of specific services such as listed in clause 5.16, based on subscription-related and invocation-related priority mechanisms. These mechanisms provide abilities such as invoking, modifying, maintaining and releasing QoS Flows with priority and delivering QoS Flow packets according to the QoS characteristics under network congestion conditions.
Subscription-related Priority Mechanisms include the ability to prioritize flows based on subscription information, including the prioritization of RRC Connection Establishment based on Unified Access Control mechanisms and the establishment of prioritized QoS Flows.
Invocation-related Priority Mechanisms include the ability for the service layer to request/invoke the activation of prioritized QoS Flows through an interaction over Rx/N5 and packet detection in the UPF.
QoS Mechanisms applied to established QoS Flows include the ability to fulfil the QoS characteristics of QoS Flows through preservation of differentiated treatment for prioritized QoS Flow and resource distribution prioritization.
Messages associated with priority services that are exchanged over service-based interfaces may include a Message Priority header to indicate priority information, as specified in TS 23.502 [3] and TS 29.500 [49].
In addition, the separation of concerns between the service classification provided by the core network through the association of Service Data Flows to QoS and the enforcing of QoS differentiation in (R)AN through the association of QoS Flows to Data Radio bearers, supports the prioritization of QoS Flows when a limitation of the available data radio bearers occurs.
In addition, it also includes the ability for the service layer to provide instructions on how to perform pre-emption of media flows with the same priority assigned through an interaction over Rx as defined in TS 23.503 [45]. |
fbecc7f0dcf9784c6066646052ab0c0e | 23.501 | 5.22.2 Subscription-related Priority Mechanisms | Subscription-related mechanisms which are always applied:
- (R)AN: During initial Access Network RRC Connection Establishment procedure or RRC Connection Resume procedure, the Establishment Cause or Resume Cause, respectively is set to indicate that special treatment is to be applied by the (R)AN in the radio resource allocation as specified in clause 5.2 for 3GPP access.
- UDM: As defined in clause 5.2.3 of TS 23.502 [3], the UE subscription data in the UDM contains an MPS subscription indication (i.e. MPS priority indication and optionally MPS for Messaging indication) and an MCX subscription indication (i.e. MCX priority) for the UE that has subscription to MPS and MCX, respectively. The MPS priority, the MPS for Messaging indication and the MCX priority, if available, are provided to the AMF via the Registration or the UE Configuration Update procedure as defined in clause 4.2 of TS 23.502 [3]. The MPS for Messaging indication sets(enables)/clears(disables) the MPS priority treatment of Messaging service (i.e. SMS over NAS, SMS over IP and messaging over IMS are all controlled by one indication) for an MPS-subscribed UE. The MPS for Messaging indication is used in communications between NFs and network entities in the CN. The MPS for Messaging indication is not delivered to UE. The MPS for Messaging indication parameter may be provisioned by an AF via NEF as described in clause 4.15.6 of TS 23.502 [3].
- AMF: Following Access Network RRC Connection Establishment procedure, the receipt of the designated Establishment Cause (i.e. high priority access) by the AMF will result in priority handling of the "Initial UE Message" received as part of the Registration procedures of clause 4.2.2 of TS 23.502 [3]. If the AMF did not receive a designated Establishment Cause (i.e. high priority access), but when the AMF determines that there is a MPS priority (or MCX priority) in the UDM for that UE, the AMF shall provide priority handling for that UE at that time and shall provide the MPS priority (or MCX priority) to the UE via the Registration or the UE Configuration Update procedure, as defined in clause 4.2 of TS 23.502 [3]. In addition, certain exemptions to Control Plane Congestion and Overload Control are provided as specified in clause 5.19. For the MPS for Messaging indication, the AMF uses the parameter to determine the Message priority header related to SMS delivery and as described in clause 4.13 of TS 23.502 [3].
Subscription-related mechanisms which are conditionally applied:
- UE: When barring control parameters are broadcast by the RAN, access barring based on Access Identity(es) configured in the USIM, or provided by the network as part of the Registration or UE Configuration Update procedure and/or an Access Category, is applied prior to an initial upstream transmission (initial or resume) for the UE which provides a mechanism to limit transmissions from UEs categorized as non-prioritized, while allowing transmissions from UEs categorized as prioritized (such as MPS subscribed UEs), during the RRC Connection Establishment procedure and RRC Connection Resume procedure as specified in clause 5.2. A UE categorized as prioritized (e.g. MPS subscribed UEs), as specified in clause 5.16.5:
- in CM-IDLE state, uses a high priority Establishment Cause in the UE Triggered Service Request procedure; and
- in RRC_INACTIVE state, uses a high priority Resume Cause in the Connection Resume procedure.
- UDM: One or more ARP priority levels are assigned for prioritized or critical services. The ARP of the prioritized QoS Flows for each DN is set to an appropriate ARP priority level. The 5QI is from the standard value range as specified in clause 5.7.2.7. In addition, Priority Level may be configured for the standardized 5QIs and if configured, it overwrites the default value specified in the QoS characteristics Table 5.7.4-1.
- PCF: If the "IMS Signalling Priority" information is set for the subscriber in the UDR and the PCF modifies the ARP of the QoS Flow used for IMS signalling, for each DN which supports prioritized services leveraging on IMS signalling, to an appropriate ARP priority level assigned for that service.
- SMF/AMF: Based on the QoS parameters (e.g. ARPs, 5QI) of a PDU Session, the SMF may provide a PDU Session Priority to the AMF. The AMF stores the information for further PDU Session related priority handling (i.e. for determination of Message Priority header of subsequent message related to the PDU Session).
NOTE: For a UE with MPS subscription, the MPS priority per DN can be different (e.g. only certain specific DNs has MPS priority while other DNs have no MPS priority). |
fbecc7f0dcf9784c6066646052ab0c0e | 23.501 | 5.22.3 Invocation-related Priority Mechanisms | The generic mechanisms used based on invocation-related Priority Mechanisms for prioritised services are based an interaction with an Application Function and between the Application Function and the PCF over Rx/N5 interface.
These mechanisms apply to mobile originated as well as mobile terminated SIP call/sessions (clause 5.21 of TS 23.228 [15]) and Priority PDU connectivity services including MPS for Data Transport Service.
NOTE 1: Clause 5.21 of TS 23.228 [15] is applicable to 5GS, with the understanding that the term PCRF corresponds to PCF in the 5GS.
Invocation-related mechanisms for Mobile Originations e.g. via SIP/IMS:
- PCF: When an indication for a session arrives over the Rx/N5 Interface and the UE does not have priority for the signalling QoS Flow, the PCF derives the ARP and 5QI parameters, plus associated QoS characteristics as appropriate, of the QoS Flow for Signalling as per Service Provider policy as specified in clause 6.1.3.11 of TS 23.503 [45].
- PCF: For sessions such as MPS, when establishing or modifying a QoS Flow for media as part of the session origination procedure, the PCF selects the ARP and 5QI parameters, plus associated QoS characteristics as appropriate, to provide priority treatment to the QoS Flow(s).
- PCF: When all active sessions to a particular DN are released and the UE is not configured for priority treatment to that particular PDU Session for a DN, the PCF will downgrade the IMS Signalling QoS Flows from appropriate settings of the ARP and 5QI parameters, plus associated QoS characteristics as appropriate, to those entitled by the UE based on subscription.
- SMF/AMF: Based on the QoS parameters (e.g. ARP, 5QI) of a PDU Session, the SMF may provide a PDU Session Priority to the AMF. The AMF stores the information for further PDU Session related priority handling (i.e. for determination of Message Priority header of subsequent message related to the PDU Session).
Invocation-related mechanisms for Mobile Terminations e.g. via SIP/IMS:
- PCF: When an indication for a session arrives over the Rx/N5 Interface, mechanisms as described above for Mobile Originations are applied.
- UPF: If an IP packet arrives at the UPF for a UE that is CM-IDLE, the UPF sends a "Data Notification" including the information to identify the QoS Flow for the DL data packet to the SMF, as specified in clause 4.2.3.3 of TS 23.502 [3].
- SMF: If a " Data Notification" message arrives at the SMF for a QoS Flow associated with an ARP priority level value that is entitled for priority use, delivery of priority indication during the Paging procedure is provided by inclusion of the ARP in the N11 interface "N1N2MessageTransfer" message, as specified in clause 4.2.3.3 of TS 23.502 [3]. Based on the QoS parameters (e.g. ARP, 5QI) of a PDU Session, the SMF may also provide a PDU Session Priority to the AMF.
- AMF: If an "N1N2MessageTransfer" message arrives at the AMF:
- from an SMF containing an ARP priority level value that is entitled for priority use;
- from the SMSF with an MT-SMS related request and either an indication that MPS for Messaging is enabled for the UE from the UDM, or with a Message Priority header value used for MPS for Messaging;
the AMF:
- handles the request with priority and includes the "Paging Priority" IE in the N2 "Paging" message set to a value assigned to indicate that there is an IP packet at the UPF or an MT-SMS at the SMSF entitled to priority treatment, as specified in clause 4.2.3.3 or clause 4.13.3.6 of TS 23.502 [3], when the UE is CM-IDLE; or
- determines the service priority ("RAN Paging Priority" IE as specified in TS 38.413 [34]) and includes the service priority in the N2 Downlink NAS Transport message to the NG-RAN, when the UE is CM-CONNECTED.
If a PDU Session Priority is provided by the SMF, the AMF stores the information for further PDU Session related priority handling (i.e. for determination of Message Priority header of subsequent message related to the PDU Session).
- SMF: For a UE that is not configured for priority treatment, upon receiving the "N7 Session Management Policy Modification" message from the PCF with an ARP priority level that is entitled for priority use, the SMF sends an "N1N2MessageTransfer" to update the ARP for the Signalling QoS Flows, as specified in clause 4.3.3.2 of TS 23.502 [3]. Based on the QoS parameters (e.g. ARP, 5QI) of a PDU Session, the SMF may also provide a PDU Session Priority to the AMF.
- AMF: Upon receiving the "N1N2MessageTransfer" message from the SMF with an ARP priority level that is entitled for priority use, the AMF updates the ARP for the Signalling QoS Flows, as specified in clause 4.3.3.2 of TS 23.502 [3].
- NG-RAN: Inclusion of the "Paging Priority" in the N2 "Paging" message or the "RAN Paging Priority" and ARP value associated with priority service (e.g. MPS) in other N2 messages (e.g. Downlink NAS Transport, PDU Session Resource Setup or Modify message) triggers priority handling of paging, including over Xn interface, in times of congestion at the NG-RAN, as specified in clauses 4.2.3.3, 4.8 and 4.13.3 of TS 23.502 [3], respectively.
Invocation-related mechanisms for the Priority PDU connectivity services:
- PCF: If the state of the Priority PDU connectivity services is modified from disabled to enabled, the QoS Flow(s) controlled by the Priority PDU connectivity services are established/modified to have the service appropriate settings of the ARP and 5QI parameters, plus associated QoS characteristics as appropriate, using the PDU Session Modification procedure as specified in clause 4.3.3 of TS 23.502 [3].
- PCF: If the state of Priority PDU connectivity services is modified from enabled to disabled, the QoS Flow(s) controlled by the Priority PDU connectivity services are modified from Priority PDU connectivity service appropriate settings of the ARP and 5QI parameters, plus associated QoS characteristics as appropriate, to those entitled by the UE as per subscription, using the PDU Session Modification procedure as specified in clause 4.3.3 of TS 23.502 [3].
- SMF/AMF: Based on the QoS parameters (e.g. ARP, 5QI) of a PDU Session, the SMF may provide a PDU Session Priority to the AMF. The AMF stores the information for further PDU Session related priority handling (i.e. for determination of Message Priority header of subsequent message related to the PDU Session).
Invocation-related mechanisms for MPS for Data Transport Service:
- MPS for Data Transport Service follows the same steps as those for Priority PDU connectivity services. The QoS Flows that will be subject to MPS for Data Transport Service are based on operator policy and regulations by means of local PCF configuration.
NOTE 2: If no configuration is provided, MPS for Data Transport Service applies to the QoS Flow associated with the default QoS rule. |
fbecc7f0dcf9784c6066646052ab0c0e | 23.501 | 5.22.4 QoS Mechanisms applied to established QoS Flows | Mechanisms applied to established QoS Flows:
- (R)AN: QoS Flows requested in the Xn "Handover Request" or N2 "Handover Request" which are marked as entitled to priority by virtue of inclusion of an ARP value from the set allocated by the Service Provider for prioritised services are given priority over requests for QoS Flows which do not include an ARP from the set as specified in clause 4.9 of TS 23.502 [3].
- SMF: Congestion management procedures in the SMF will provide priority to QoS Flows established for sessions during periods of extreme overload. Prioritised services are exempt from any session management congestion controls. See clause 5.19.
- AMF: Congestion management procedures in the AMF will provide priority to any Mobility Management procedures required for the prioritised services during periods of extreme overload. Prioritised services are exempt from any Mobility Management congestion controls. See clause 5.19.5.
- QoS Flows whose ARP parameter is from the set allocated by the Service Provider for prioritised services' use shall be exempt from release during QoS Flow load rebalancing.
- (R)AN, UPF: IMS Signalling Packets associated with prioritised services' use are handled with priority. Specifically, during times of severe congestion when it is necessary to drop packets on the IMS Signalling QoS Flow, or QoS Flow supporting MPS for Data Transport Service signalling, to ensure network stability, these FEs shall drop packets not associated with priority signalling such as MPS or Mission Critical services before packets associated with priority signalling. See clauses 5.16.5 and 5.16.6.
- (R)AN, UPF: During times of severe congestion when it is necessary to drop packets on a media QoS Flow to ensure network stability, these FEs shall drop packets not associated with priority sessions such as MPS or Mission Critical services before packets associated with priority sessions. See clauses 5.16.5 and 5.16.6. |
fbecc7f0dcf9784c6066646052ab0c0e | 23.501 | 5.23 Supporting for Asynchronous Type Communication | Asynchronous type communication (ATC) enables 5GC to delay synchronizing UE context with the UE, so as to achieve an efficient signalling overhead and increase system capacity. The support of ATC is optional for the AMF.
5GC supports asynchronous type communication with the following functionality:
- Capability to store the UE context based on the received message and synchronize the UE context with the involved network functions or UE later;
For network function (e.g. PCF, UDM, etc.) triggered signalling procedure (e.g. network triggered Service Request procedure, network triggered PDU Session Modification procedure, etc.), if the UE CM state in the AMF is CM-IDLE state and the requesting network function indicates to the AMF the ATC is allowed for the signalling, if the AMF supports the ATC feature, the AMF may update and store the UE context based on the received message without paging UE immediately. When the UE CM state in the AMF enters CM-CONNECTED state, the AMF forwards N1 and N2 message to synchronize the UE context with the (R)AN and/or the UE.
If the originating NF does not require immediate delivery, it may indicate that the AMF is allowed to use ATC.
NOTE: Pre-Rel-17 AMF implementation cannot decode the indication that ATC is allowed. |
fbecc7f0dcf9784c6066646052ab0c0e | 23.501 | 5.24 3GPP PS Data Off | This feature, when activated by the user, prevents traffic via 3GPP access of all IP packets, Unstructured and Ethernet data except for those related to 3GPP PS Data Off Exempt Services. The 3GPP PS Data Off Exempt Services are a set of operator services, defined in TS 22.011 [25] and TS 23.221 [23], that are the only allowed services when the 3GPP PS Data Off feature has been activated by the user. The 5GC shall support 3GPP PS Data Off operation in both non-roaming and roaming scenarios.
UEs may be configured with up to two lists of 3GPP PS Data Off Exempt Services and the list(s) are provided to the UEs by HPLMN via Device Management or UICC provisioning. When the UE is configured with two lists, one list is valid for the UEs camping in the home PLMN and the other list is valid for any VPLMN the UE is roaming in. When the UE is configured with a single list, without an indication to which PLMNs the list is applicable, then this list is valid for the home PLMN and any PLMN the UE is roaming in.
NOTE 1: The operator needs to ensure coordinated list(s) of 3GPP Data Off Exempt Services provisioned in the UE and configured in the network.
The UE reports its 3GPP PS Data Off status in PCO (Protocol Configuration Option) to (H-)SMF during UE requested PDU Session Establishment procedure for establishment of a PDU Session associated with 3GPP access and/or non-3GPP access. The UE does not need to report PS Data Off status during the PDU Session Establishment procedure for handover of the PDU Session between 3GPP access and non 3GPP access if 3GPP PS Data Off status is not changed since the last report. The PS Data Off status for a PDU Session does not affect data transfer over non-3GPP access.
If 3GPP PS Data Off is activated, the UE prevents the sending of uplink IP packets, Unstructured and Ethernet data except for those related to 3GPP PS Data Off Exempt Services, based on the pre-configured list(s) of Data Off Exempt Services.
If 3GPP PS Data Off is activated for a UE with MA PDU Sessions established through the ATSSS feature (see clause 5.32), the data transferred over the non-3GPP access of the MA PDU sessions are unaffected, which is ensured by the policy for ATSSS Control as specified in clause 5.32.3.
The UE shall immediately report a change of its 3GPP PS Data Off status in PCO by using UE requested PDU Session Modification procedure. This also applies to the scenario of inter-RAT mobility to NG-RAN and to scenarios where the 3GPP PS Data Off status is changed when the session management back-off timer is running as specified in clause 5.19.7.3 and clause 5.19.7.4. For UEs in Non-Allowed Area (or not in Allowed Area) as specified in clause 5.3.4.1, the UE shall also immediately report a change of its 3GPP PS Data Off status for the PDU Session. For UEs moving out of LADN area and the PDU Session is still maintained as specified in clause 5.6.5, the UE shall also immediately report a change of its 3GPP PS Data Off status for the PDU Session.
The additional behaviour of the SMF for 3GPP PS Data Off is controlled by local configuration or policy from the PCF as defined in TS 23.503 [45].
NOTE 2: For the PDU Session used for IMS services, the 3GPP Data Off Exempt Services are enforced in the IMS domain as specified TS 23.228 [15]. Policies configured in the (H-)SMF/PCF need to ensure those services are always allowed when the 3GPP Data Off status of the UE is set to "activated". |
fbecc7f0dcf9784c6066646052ab0c0e | 23.501 | 5.25 Support of OAM Features | |
fbecc7f0dcf9784c6066646052ab0c0e | 23.501 | 5.25.1 Support of Tracing: Signalling Based Activation/Deactivation of Tracing | 5GS supports tracing as described in TS 32.421 [66]. 5GS support may include subscriber tracing (tracing targeting a SUPI) or equipment tracing (tracing targeting a PEI) but also other forms of tracing further described in TS 32.421 [66] and TS 32.422 [192].
NOTE 1: TS 23.501 / TS 23.502 [3] / TS 23.503 [45] only describe how 5GS signalling supports delivery of Trace Requirements about a UE (Signalling Based Activation/Deactivation of Tracing). OAM delivery of tracing requirements as well as the transfer of tracing results to one or more Operations Systems are out of scope of these documents.
The content of Trace Requirements about a UE (e.g. trace reference, address of the Trace Collection Entity, etc.) is defined in TS 32.421 [66].
Trace Requirements about a UE may be configured in subscription data of the UE and delivered together with other subscription data by the UDM towards the AMF, the SMF and/or the SMSF.
The AMF propagates Trace Requirements about a UE received from the UDM to network entities not retrieving subscription information from UDM, i.e. to the 5G-AN, to the AUSF and to the PCF. The AMF also propagates Trace Requirements to the SMF and to the SMSF. If the I-SMF or V-SMF is needed for the PDU session, the AMF propagates Trace Requirements to the I-SMF or V-SMF. The I-SMF or V-SMF also propagates Trace Requirements received from the AMF to the I-UPF or V-UPF (over N4).
Trace Requirements about a UE may be sent by the AMF to the 5G-AN as part of:
- the N2 procedures used to move the UE from CM-IDLE to CM-CONNECTED or,
- the N2 procedures to request a Hand-over from a target NG-RAN or,
- a stand-alone dedicated N2 procedure when tracing is activated while the UE is CM-CONNECTED.
Trace Requirements about a UE sent to a 5G-AN shall not contain information on the SUPI or on the PEI of the UE. Trace Requirements are directly sent from Source to Target NG-RAN in the case of Xn Hand-Over.
The SMF propagates Trace Requirements about a UE received from the UDM to the UPF (over N4) and to the PCF. The SMF provides Trace Requirements to the PCF when it has selected a different PCF than the one received from the AMF.
Once the SMF or the SMSF has received subscription data, Trace Requirements received from UDM supersede Trace requirements received from the AMF. Trace Requirements are exchanged on N26 between the AMF and the MME. |
fbecc7f0dcf9784c6066646052ab0c0e | 23.501 | 5.25.2 Support of OAM-based 5G VN group management | 5GS supports 5G LAN-type service as defined in clause 5.29. 5G LAN-type service includes the 5G VN group management that can be configured by a network administrator.
The parameters for 5G VN group is defined in clause 5.29.
The 5G VN group parameters about a UE may be configured in subscription data of the UE and delivered together with other subscription data by the UDM towards the AMF and SMF. |
fbecc7f0dcf9784c6066646052ab0c0e | 23.501 | 5.25.3 Signalling Based Activation of QoE Measurement Collection | 5GS may support QoE Measurement Collection (QMC) as described in TS 28.405 [190]. 5GS may support Signalling based activation and Management based activation as specified in TS 38.300 [27] and TS 28.405 [190].
To enable signalling based activation in NR, the QMC Configuration information is configured in subscription data of the UE. 5GC provides the QMC Configuration information to the NG-RAN as described in TS 23.502 [3]. |
fbecc7f0dcf9784c6066646052ab0c0e | 23.501 | 5.26 Configuration Transfer Procedure | The purpose of the Configuration Transfer is to enable the transfer of information between two RAN nodes at any time via NG interface and the Core Network. An example of application is to exchange the RAN node's IP addresses in order to be able to use Xn interface between the NG-RAN node for Self-Optimised Networks (SON), as specified in TS 38.413 [34]. |
fbecc7f0dcf9784c6066646052ab0c0e | 23.501 | 5.26.1 Architecture Principles for Configuration Transfer | Configuration Transfer between two RAN node provides a generic mechanism for the exchange of information between applications belonging to the RAN nodes.
In order to make the information transparent for the Core Network, the information is included in a transparent container that includes source and target RAN node addresses, which allows the Core Network nodes to route the messages. The mechanism is depicted in figure 5.26 1.
Figure 5.26-1: inter NG-RAN Configuration Transfer basic network architecture
The NG-RAN transparent containers are transferred from the source NG-RAN node to the destination NG-RAN node by use of Configuration Transfer messages.
A Configuration Transfer message is used from the NG-RAN node to the AMF over N2 interface, a AMF Configuration Transfer message is used from the AMF to the NG-RAN over N2 interface and a Configuration Transfer Tunnel message is used to tunnel the transparent container from a source AMF to a target AMF over the N14 interface.
Each Configuration Transfer message carrying the transparent container is routed and relayed independently by the core network node(s). |
fbecc7f0dcf9784c6066646052ab0c0e | 23.501 | 5.26.2 Addressing, routing and relaying | |
fbecc7f0dcf9784c6066646052ab0c0e | 23.501 | 5.26.2.1 Addressing | All the Configuration Transfer messages contain the addresses of the source and destination RAN nodes. An NG-RAN node is addressed by the Target NG-RAN node identifier. |
fbecc7f0dcf9784c6066646052ab0c0e | 23.501 | 5.26.2.2 Routing | The following description applies to all the Configuration Transfer messages used for the exchange of the transparent container.
The source RAN node sends a message to its core network node including the source and destination addresses. The AMF uses the destination address to route the message to the correct AMF via the N14 interface.
The AMF connected to the destination RAN node decides which RAN node to send the message to, based on the destination address. |
fbecc7f0dcf9784c6066646052ab0c0e | 23.501 | 5.26.2.3 Relaying | The AMF performs relaying between N2 and N14 messages as described in TS 38.413 [34], TS 29.518 [71]. |
fbecc7f0dcf9784c6066646052ab0c0e | 23.501 | 5.27 Enablers for Time Sensitive Communications, Time Synchronization and Deterministic Networking | |
fbecc7f0dcf9784c6066646052ab0c0e | 23.501 | 5.27.0 General | This clause describes 5G System features that can be used independently or in combination to enable time-sensitive communication, time synchronization and deterministic networking:
- Delay-critical GBR;
- A hold and forward mechanism to schedule traffic as defined in IEEE Std 802.1Q [98] for Ethernet PDU Sessions in DS-TT and NW-TT (see clause 5.27.4) to de-jitter flows that have traversed the 5G System if the 5G System is to participate transparently as a bridge in a TSN network;
- TSC Assistance Information: describes TSC flow traffic characteristics as described in clause 5.27.2 that may be provided optionally for use by the gNB, to allow more efficiently schedule radio resources for periodic traffic and applies to PDU Session type Ethernet and IP.
- Time Synchronization: describes how 5GS can operate as a PTP Relay (IEEE Std 802.1AS [104]), as a Boundary Clock or as Transparent Clock (IEEE Std 1588 [126]) for PDU Session type Ethernet and IP and how 5GS can detect and report the status of the time synchronization.
- RAN feedback for BAT offset and adjusted periodicity describes a mechanism supported by NG-RAN and 5G CN that enables AF to adapt to received BAT offset and adjusted periodicity from NG-RAN for a given traffic flow.
The 5G System integration as a bridge in an IEEE 802.1 TSN network as described in clause 5.28 can make use of all features listed above.
To support any of the above features to enable time-sensitive communication, time synchronization and deterministic networking, during the PDU Session establishment, the UE shall request to establish a PDU Session as an always-on PDU Session and the PDU Sessions are established as Always-on PDU session as described in clause 5.6.13. In this release of the specification, to use any of the above features to enable time-sensitive communication, time synchronization and deterministic networking:
- Home Routed PDU Sessions are not supported;
- PDU Sessions are supported only for SSC mode 1;
- Service continuity is not supported when the UE moves from 5GS to EPS .i.e. interworking with EPS is not supported for a PDU Session for time synchronization or TSC or deterministic networking. |
fbecc7f0dcf9784c6066646052ab0c0e | 23.501 | 5.27.1 Time Synchronization | |
fbecc7f0dcf9784c6066646052ab0c0e | 23.501 | 5.27.1.1 General | For supporting time synchronization service, the 5GS is configured to operate in one or multiple PTP instances and to operate in one of the following modes (if supported) for each PTP instance:
1) as time-aware system as described in IEEE Std 802.1AS [104],
2) as Boundary Clock as described in IEEE Std 1588 [126], provisioned by the profiles supported by this 3GPP specification including SMPTE Profile for Use of IEEE Std 1588 [126] Precision Time Protocol in Professional Broadcast Applications ST 2059-2:2015 [127];
NOTE 1: Via proper configuration of the IEEE Std 1588 [126] data set members, the 5G internal system clock can become the time source for the PTP grandmaster function for the connected networks in the case of mode 1 and mode 2.
NOTE 2: In some cases where the 5G internal system clock is the time source for the PTP grandmaster function for the connected networks, it might not be required for the UE to receive gPTP or PTP messages over user plane. The UE and DS-TT uses the 5G timing information and generates the necessary gPTP or PTP message for the end station, if needed (this is implementation specific).
3) as peer-to-peer Transparent Clock as described in IEEE Std 1588 [126], provisioned by the profiles supported by this 3GPP specification including SMPTE Profile for Use of IEEE Std 1588 Precision Time Protocol in Professional Broadcast Applications ST 2059-2:2015 [127]; or
4) as end-to-end Transparent Clock as described in IEEE Std 1588 [126], provisioned by the profiles supported by this 3GPP specification including SMPTE Profile for Use of IEEE Std 1588 Precision Time Protocol in Professional Broadcast Applications ST 2059-2:2015 [127].
NOTE 3: When the GM is external, the operation of 5GS as Boundary Clock assumes that profiles that are supported by the 5GS allows the exemption specified in clauses 9.5.9 and 9.5.10 of IEEE Std 1588 [126] where the originTimestamp (or preciseOriginTimestamp in case of two-step operation) is not required to be updated with the syncEventEgressTimestamp (and a Local PTP Clock locked to the external GM). As described in clause 5.27.1.2.2, only correctionField is updated with the 5GS residence time and link delay, in a similar operation as specified by IEEE Std 802.1AS [104].
The configuration of the time synchronization service in 5GS for option 1 by TSN AF and CNC is described in clause 5.28.3 and for options 1-4 by AF/NEF and TSCTSF in clause 5.27.1.8 and clause 5.28.3.
The 5GS shall be modelled as an IEEE Std 802.1AS [104] or IEEE Std 1588 [126] compliant entity based on the above configuration.
NOTE 4: This release of the specification does not support the PTP management mechanism or PTP management messages as described in clause 15 in IEEE Std 1588 [126].
The DS-TT and NW-TT at the edge of the 5G system may support the IEEE Std 802.1AS [104] or other IEEE Std 1588 [126] profiles' operations respective to the configured mode of operation. The UE, gNB, UPF, NW-TT and DS- TTs are synchronized with the 5G GM (i.e. the 5G internal system clock) which shall serve to keep these network elements synchronized. The TTs located at the edge of 5G system fulfil some functions related to IEEE Std 802.1AS [104] and may fulfil some functions related to IEEE Std 1588 [126], e.g. (g)PTP support and timestamping. Figure 5.27.1-1 illustrates the 5G and PTP grandmaster (GM) clock distribution model via 5GS.
Figure 5.27.1-1: 5G system is modelled as PTP instance for supporting time synchronization
Figure 5.27.1-1 depicts the two synchronizations systems considered: the 5G Clock synchronization and the (g)PTP domain synchronization.
- 5G Access Stratum-based Time Distribution: Used for NG RAN synchronization and also distributed to the UE. The 5G Access Stratum-based Time Distribution over the radio interface towards the UE is specified in TS 38.331 [28]. This method may be used to either further distribute the 5G timing to devices connected to a UE (using implementation-specific means) or to support the operation of the (g)PTP-based time distribution method.
- (g)PTP-based Time Distribution: Provides timing among entities in a (g)PTP domain. This process follows the applicable profiles of IEEE Std 802.1AS [104] or IEEE Std 1588 [126]. This method relies on the 5G access stratum-based time distribution method to synchronize the UE/DS-TT and on the 5GS time synchronization to synchronize the gNB (which, in turn, may synchronize the DS-TT) and the NW-TT.
The gNB needs to be synchronized to the 5G GM clock.
The 5GS supports two methods for determining the grandmaster PTP Instance and the time-synchronization spanning tree.
- Method a), BMCA procedure.
- Method b), local configuration.
This is further described in clause 5.27.1.6. |
fbecc7f0dcf9784c6066646052ab0c0e | 23.501 | 5.27.1.2 Distribution of timing information | |
fbecc7f0dcf9784c6066646052ab0c0e | 23.501 | 5.27.1.2.1 Distribution of 5G internal system clock | The 5G internal system clock shall be made available to all user plane nodes in the 5G system. The UPF and NW-TT may get the 5G internal system clock via the underlying PTP compatible transport network with mechanisms outside the scope of 3GPP. The 5G internal system clock shall be made available to UE with signalling of time information related to absolute timing of radio frames as described in TS 38.331 [28]. The 5G internal system clock shall be made available to DS-TT by the UE. |
fbecc7f0dcf9784c6066646052ab0c0e | 23.501 | 5.27.1.2.2 Distribution of grandmaster clock and time-stamping | 5.27.1.2.2.1 Distribution of gPTP Sync and Follow_Up messages
The mechanisms for distribution of TSN GM clock and time-stamping described in this clause are according to IEEE Std 802.1AS [104].
NOTE 1: It means Externally-observable behaviour of the 5GS bridge needs to comply with IEEE Std 802.1AS [104].
For downlink Time Synchronization, upon reception of a downlink gPTP message from NW-TT port in Follower state, the NW-TT makes an ingress timestamping (TSi) for each gPTP event (Sync) message and uses the cumulative rateRatio received inside the gPTP message payload (carried within Sync message for one-step operation or Follow_up message for two-step operation) to calculate the link delay from the upstream TSN node (gPTP entity connected to NW-TT) expressed in TSN GM time as specified in IEEE Std 802.1AS [104]. NW-TT then calculates the new cumulative rateRatio (i.e. the cumulative rateRatio of the 5GS) as specified in IEEE Std 802.1AS [104] and modifies the gPTP message payload (carried within Sync message for one-step operation or Follow_up message for two-step operation) as follows:
- Adds the link delay from the upstream TSN node in TSN GM time to the correction field.
- Replaces the cumulative rateRatio received from the upstream TSN node with the new cumulative rateRatio.
- Adds TSi in the Suffix field of the gPTP packet as described in clause H.2.
The UPF/NW-TT uses the ingress port number of the NW-TT and domainNumber and sdoId in the received gPTP message to assign the gPTP message to a PTP instance in the NW-TT. If the NW-TT does not have a matching PTP instance, the UPF/NW-TT discards the message. The UPF/NW-TT then forwards the gPTP message from TSN network to the PTP ports in DS-TT(s) in Leader state within this PTP instance via PDU sessions terminating in this UPF that the UEs have established to the TSN network. The UPF/NW-TT also forwards the gPTP message to the PTP ports in NW-TT in Leader state within this PTP instance. All gPTP messages are transmitted on a QoS Flow that complies with the residence time upper bound requirement specified in IEEE Std 802.1AS [104].
NOTE 2: Leader and Follower terms in this specification maps to Master and Slave terms respectively for (g)PTP time synchronization as specified in IEEE Std 802.1AS [104] and IEEE Std 1588 [126]. This terminology can require update depending on the IEEE 1588 WG response to SA WG2.
NOTE 3: The sum of the UE-DS-TT residence time and the PDB of the QoS Flow needs to be lower than the residence time upper bound requirement for a time-aware system specified in IEEE Std 802.1AS [104] in the following cases:
a) If the PTP port in DS-TT is in Follower state and a PTP port in the NW-TT is in Leader state; or
b) a PTP port in DS-TT is in Leader state and a PTP port in NW-TT is in Follower state.
NOTE 4: If the PTP port in DS-TT is in a Follower state and a PTP port in another DS-TT is in Leader state, then the sum of the residence time for these two DS-TT ports and the PDB of the QoS flow of the two PDU Sessions needs to be lower than the residence time upper bound requirement for a time-aware system specified in IEEE Std 802.1AS [104].
A UE receives the gPTP messages and forwards them to the DS-TT. The DS-TT then creates egress timestamping (TSe) for the gPTP event (Sync) messages for external TSN working domains. The difference between TSi and TSe is considered as the calculated residence time spent within the 5G system for this gPTP message expressed in 5GS time. The DS-TT then uses the rateRatio contained inside the gPTP message payload (carried within Sync message for one-step operation or Follow_up message for two-step operation) to convert the residence time spent within the 5GS in TSN GM time and modifies the payload of the gPTP message that it sends towards the downstream TSN node (gPTP entity connected to DS-TT) as follows:
- Adds the calculated residence time expressed in TSN GM time to the correction field.
- Removes Suffix field that contains TSi.
If the ingress DS-TT has indicated support of the IEEE Std 802.1AS [104] PTP profile as described in clause K.2.1 and the network has configured a PTP instance with the IEEE Std 802.1AS [104] PTP profile for the ingress DS-TT, the ingress DS-TT performs the following operations for received UL gPTP messages for the PTP instance:
- Adds the link delay from the upstream TSN node (gPTP entity connected to DS-TT) in TSN GM time to the correction field.
- Replaces the cumulative rateRatio received from the upstream TSN node (gPTP entity connected to DS-TT) with the new cumulative rateRatio.
- Adds TSi in the Suffix field of the gPTP packet.
The UE transparently forwards the gPTP message from DS-TT to the UPF/NW-TT. If the ingress DS-TT port is in Passive state, the UPF/NW-TT discards the gPTP messages. If the ingress DS-TT port is in Follower state, the UPF/NW-TT forwards the gPTP messages as follows:
- In the case of synchronizing end stations behind NW-TT, the egress port is in UPF/NW-TT. For the received UL gPTP messages, the egress UPF/NW-TT performs the following actions:
- Adds the calculated residence time expressed in TSN GM to the correction field.
- Removes Suffix field that contains TSi.
- In the case of synchronizing TSN end stations behind DS-TT, the egress TT is DS-TT of the other UE and the UPF/NW-TT uses the port number of the ingress DS-TT and domainNumber and sdoId in the received gPTP message to assign the gPTP message to a PTP instance in the NW-TT. If the NW-TT does not have a matching PTP instance, the UPF/NW-TT discards the message. The UPF/NW-TT then forwards the received UL gPTP message to the PTP ports in DS-TT(s) in Leader state within this PTP instance. The egress DS-TT performs same actions as egress UPF/NW-TT in previous case.
5.27.1.2.2.2 Distribution of PTP Sync and Follow_Up messages
This clause applies if DS-TT and NW-TT support distribution of PTP Sync and Follow_Up messages. PTP support by DS-TT and NW-TT may be determined as described in clause K.2.1.
The mechanisms for distribution of PTP GM clock and time-stamping described in this clause are according to IEEE Std 1588 [126] for Transparent clock and for the case of Boundary clock when the GM is external, where the originTimestamp (or preciseOriginTimestamp) is not updated by the 5GS as described by the exemption in clause 5.27.1.1. If the 5GS acts as the GM with a PTP instance type Boundary clock, then the 5GS updates the originTimestamp (or preciseOriginTimestamp in case of two-step operation) with the 5GS internal clock, as described in clause 5.27.1.7.
NOTE 1: This means externally-observable behaviour of the PTP instance in 5GS needs to comply with IEEE Std 1588 [126].
Upon reception of a PTP event message from the upstream PTP instance, the ingress TT (i.e. NW-TT or DS-TT) makes an ingress timestamping (TSi) for each PTP event (i.e. Sync) message.
The PTP port in the ingress TT measures the link delay from the upstream PTP instance as described in clause H.4.
The PTP port in the ingress TT modifies the PTP message payload (carried within Sync message for one-step operation or Follow_Up message for two-step operation) as follows:
- (if the PTP port in the ingress TT has measured the link delay) Adds the measured link delay from the upstream PTP instance in PTP GM time to the correction field.
- (if the PTP port in the ingress TT has measured the link delay and rateRatio is used) Replaces the cumulative rateRatio received from the upstream PTP instance with the new cumulative rateRatio.
- Adds TSi in the Suffix field of the PTP message as described in clause H.2.
NOTE 2: If the 5GS is configured to use the Cumulative frequency transfer method for synchronizing clocks as described in clause 16.10 in IEEE Std 1588 [126], i.e. when the cumulative rateRatio is measured, then the PTP port in the ingress TT uses the cumulative rateRatio received inside the PTP message payload (carried within Sync message for one-step operation or Follow_Up message for two-step operation) to correct the measured link delay to be expressed in PTP GM time as specified in IEEE Std 1588 [126]. The PTP port in the ingress TT then calculates the new cumulative rateRatio (i.e. the cumulative rateRatio of the 5GS) as specified in IEEE Std 1588 [126].
NOTE 3: If 5GS acts as an end-to-end Transparent Clock, since the end-to-end Transparent Clock does not support peer-to-peer delay mechanism, the residence time can be calculated with the residence time spent within the 5GS in 5G GM time and if needed, with a correction factor, for instance, as specified in Equation (6) of clause 12.2.2 of IEEE Std 1588 [126], this gives a residence time expressed in PTP GM time that is used to update the correction field of the received PTP Sync or Follow_Up message.
The PTP port in the ingress TT then forwards the PTP message to the UPF/NW-TT. The UPF/NW-TT further distributes the PTP message as follows:
- If the 5GS is configured to operate as Boundary Clock as described in IEEE Std 1588 [126], the UPF/NW-TT uses the port number of the ingress DS-TT and domainNumber and sdoId in the received PTP message to assign the PTP message to a PTP instance in the NW-TT. If the NW-TT does not have a matching PTP instance, the UPF/NW-TT discards the message. The UPF/NW-TT then regenerates the Sync and Follow_Up (for two-step operation) messages based on the received Sync and Follow_Up messages for the PTP ports in Leader state in NW-TT and DS-TT(s) within this PTP instance. The NW-TT/UPF forwards the regenerated Sync and Follow_Up (for two-step operation) messages to the Leader ports in NW-TT and the PDU session(s) related to the Leader ports in the DS-TT(s) within this PTP instance.
- If the 5GS is configured to operate as a Transparent Clock as described in IEEE Std 1588 [126], the UPF/NW-TT uses the port number of the ingress TT and domainNumber and sdoId in the received PTP message to assign the PTP message to a PTP instance in the NW-TT. If the NW-TT does not have a matching PTP instance, the UPF/NW-TT discards the message. The UPF/NW-TT then forwards the received Sync messages to PTP ports in DS-TT(s) within this PTP instance via corresponding PDU Sessions terminating to this UPF and to NW-TT ports within this PTP instance, except toward the ingress PTP port in the ingress TT.
NOTE 4: If 5GS acts as a Transparent Clock, the 5GS does not maintain the PTP port states; the ingress PTP messages received on a PTP Port are retransmitted on all other PTP Ports of the Transparent Clock subject to the rules of the underlying transport protocol.
NOTE 5: Due to the exemption described in clause 5.27.1.1, when the PTP instance in 5GS is configured to operate as a Boundary Clock, the 5GS does not need to synchronize its Local PTP Clock to the external PTP grandmaster. The PTP instance in 5GS measures the link delay and residence time and communicates these in a correction field. The externally observable behaviour of 5GS still conforms to the specifications for a Boundary Clock as described in IEEE Std 1588 [126].
The PTP port in the egress TT then creates egress timestamping (TSe) for the PTP event (i.e. Sync) messages for external PTP network. The difference between TSi and TSe is considered as the calculated residence time spent within the 5G system for this PTP message expressed in 5GS time.
The PTP port in the egress TT then uses the rateRatio contained inside the PTP message payload (if available, carried within Sync message for one-step operation or Follow_Up message for two-step operation) to convert the residence time spent within the 5GS in PTP GM time.
The PTP port in the egress TT modifies the payload of the PTP message (Sync message for one-step operation or Follow_Up message for two-step operation) that it sends towards the downstream PTP instance as follows:
- Adds the calculated residence time to the correction field.
- Removes Suffix field of the PTP message that contains TSi.
NOTE 6: If 5GS acts as an end-to-end Transparent Clock, since the end-to-end Transparent Clock does not support peer-to-peer delay mechanism, the residence time is calculated with the residence time spent within the 5GS in 5G GM time and, if needed, corrected for instance with the factor as specified in Equation (6) of clause 12.2.2 of IEEE Std 1588 [126] to get it expressed in PTP GM time. The residence time is used to update the correction field of the received PTP event (e.g. Sync or Follow_Up) message. |
fbecc7f0dcf9784c6066646052ab0c0e | 23.501 | 5.27.1.3 Support for multiple (g)PTP domains | This clause describes support for multiple domains for gPTP and PTP and for GM clocks connected to DS-TT and NW-TT and only applies if DS-TT and NW-TT support the related functionality. PTP support and support of gPTP for GM clocks connected to DS-TT by DS-TT and NW-TT may be determined as described in clause K.2.1.
Each (g)PTP domain sends its own (g)PTP messages. The (g)PTP message carries a specific PTP "domainNumber" that indicates the time domain they are referring to. The PTP port in ingress TT makes ingress timestamping (TSi) for the (g)PTP event messages of all domains and forwards the (g)PTP messages of all domains to the UPF/NW-TT that further distributes the (g)PTP messages to the egress TTs as specified in clause 5.27.1.2.2.
The PTP port in the egress TT receives the original PTP GM clock timing information and the corresponding TSi via (g)PTP messages for one or more (g)PTP domains. The PTP port in the egress TT then makes egress timestamping (TSe) for the (g)PTP event messages for every (g)PTP domain. Ingress and egress time stamping are based on the 5G system clock at NW-TT and DS-TT.
NOTE 1: An end-station can select PTP timing information of interest based on the "domainNumber" in the (g)PTP message.
The process described in clause 5.27.1.2.2 is thus repeated for each (g)PTP domain between a DS-TT and the NW-TT it is connected to.
NOTE 2: If all (g)PTP domains can be made synchronous and the synchronization can be provided by the 5G clock, the NW-TT or DS-TT(s) generates the (g)PTP event messages of all domains using 5G clock as described in clause 5.27.1.7.
NOTE 3: This Release of the specification supports multiple gPTP domains as defined in IEEE Std 802.1AS [104]. If a 5GS TSN bridge supports stream gates and/or transmission gates as defined in IEEE Std 802.1Q [98], then they operate based on a single given gPTP domain. |
fbecc7f0dcf9784c6066646052ab0c0e | 23.501 | 5.27.1.4 DS-TT and NW-TT Time Synchronization functionality | This clause describes the support of Time Synchronization functionality supported by the 5G System. Synchronization between UPF/NW-TT and NG-RAN is outside scope of 3GPP.
DS-TT and NW-TT may support the following PTP instance types:
- Boundary Clock as defined in IEEE Std 1588 [126] as described in clause 5.27.1.1;
- End-to-End Transparent Clock as defined in IEEE Std 1588 [126] as described in clause 5.27.1.1;
- Peer-to-Peer Transparent Clock as defined in IEEE Std 1588 [126] as described in clause 5.27.1.1;
- PTP Relay instance as defined in IEEE Std 802.1AS [104].
Editor's note: Support for external networks operating with IEEE Std 1588-2008 [107] is for further study.
DS-TT and NW-TT may support the following transports for PTP:
- IPv4 as defined in IEEE Std 1588 [126] Annex C;
- IPv6 as defined in IEEE Std 1588 [126] Annex D;
IEEE Std 802.3 [131] (Ethernet) as defined in IEEE Std 1588 [126] Annex E.
For operation as a Boundary clock or as a Transparent Clock, DS-TT and NW-TT may support the following path and link delay measurement methods:
- Delay request-response mechanism as described in clause 11.3 of IEEE Std 1588 [126];
- Peer-to-peer delay mechanism as defined in clause 11.4 of IEEE Std 1588 [126].
DS-TT and NW-TT may support acting as a PTP grandmaster, i.e. may support generating (g)PTP Announce, Sync and Follow_Up messages. DS-TT and NW-TT supporting (g)PTP shall support one or more PTP profiles as described in clause 20.3 of IEEE Std 1588 [126], i.e.:
- Default PTP Profiles in IEEE Std 1588 [126], Annex I;
- IEEE Std 802.1AS [104] PTP profile for transport of timing as defined in IEEE Std 802.1AS [104] Annex F;
- SMPTE Profile for Use of IEEE Std 1588 [126] Precision Time Protocol in Professional Broadcast Applications ST 2059-2:2015 [127].
TSN AF and TSCTSF may determine the PTP functionalities supported by DS-TT and NW-TT and may configure PTP instances in DS-TT and NW-TT using port and user plane node management information exchange as described in Annex K, clause K.2.
NOTE: How the TSN AF or TSCTSF assigns NW-TT port(s) of one NW-TT to different PTP instances is up to implementation. |
fbecc7f0dcf9784c6066646052ab0c0e | 23.501 | 5.27.1.5 Detection of (g)PTP Sync and Announce timeouts | The procedure described in this clause is applicable when the PTP instance in 5GS is configured to operate as a time-aware system or as a Boundary Clock and the PTP grandmaster is external to the 5GS and the BMCA procedure (Method a) is used as described in clause 5.27.1.6.
The NW-TT processes Announce messages according to IEEE Std 1588 [126].
In particular, the NW-TT shall compute and maintain the time when the Announce and Sync timeout events occur for the PTP port in a Follower state. When the 5GS is configured to operate as a time-aware system, the NW-TT shall determine the Sync and Announce message interval for the PTP Port at the other end of the link to which the Follower PTP Port in 5GS is attached, as described in IEEE Std 802.1AS [104]. When the 5GS is configured to operate as a Boundary Clock, the NW-TT shall determine Announce interval based on the configuration of the Follower port in 5GS, as described in IEEE Std 1588 [126].
The configuration of PTP instances in DS-TT and NW-TT for Sync and Announce timeouts is described in clause K.2. Upon detection of the Sync or Announce timeout event, the NW-TT shall re-evaluate the DS-TT and NW-TT port states as described in clause 5.27.1.6. |
fbecc7f0dcf9784c6066646052ab0c0e | 23.501 | 5.27.1.6 Distribution of Announce messages and best master clock selection | The procedure described in this clause is applicable if DS-TT and NW-TT support operating as a Boundary Clock described in IEEE Std 1588 [126] or as a time-aware system (support of the IEEE 802.1AS [104] PTP profile) and when the PTP instance in 5GS is configured to operate as a time-aware system or as a Boundary Clock. Whether DS-TT/NW-TT support operating as a Boundary Clock or as a time-aware system may be determined as described in clause K.2.1.
The externally-observable behaviour of the Announce message handling by 5GS needs to comply with IEEE Std 802.1AS [104] or IEEE Std 1588 [126], respective to the configured mode of operation.
The DS-TT forwards the received Announce messages to NW-TT over User plane. The NW-TT port forwards the received Announce messages from N6 interface to NW-TT.
The NW-TT maintains the PTP port state for each DS-TT port and NW-TT port. The PTP port states may be determined by NW-TT either via:
- Method a), BMCA procedure.
- Method b), local configuration.
When Method b) is used, the following applies:
- When the PTP GM is external to the 5GS, for one of the NW-TT or DS-TT ports (per each PTP domain) the PTP port state is Follower and for all other NW-TT and DS-TT ports of the same PTP domain the PTP port state is set either to Passive or Leader (depending on implementation).
- When the 5GS is configured as a grandmaster for a (g)PTP domain for the connected networks, all NW-TT ports and DS-TT ports are set to Leader state for that (g)PTP domain.
The local configuration of PTP port states in DS-TT and NW-TT for Method b is described in clause K.2.
When the Method a) is used (PTP port states are determined by BMCA procedure), the NW-TT needs to process the received Announce messages (from NW-TT port(s) and over user plane from the DS-TT(s)) for BMCA procedure, determine port states within the 5GS and maintain the Leader-Follower hierarchy.
The DS-TT maintains the Disabled, Initializing and Faulty (if applicable) state for the PTP ports in DS-TT. While in Disabled, Initializing or Faulty state, the port in the DS-TT discards any (g)PTP messages it may receive from the upstream PTP instance or from the NW-TT via user plane.
When the 5GS Clock is determined as a grandmaster for a (g)PTP domain, the Announce messages are distributed as described in clause 5.27.1.7.
When the grandmaster is external to the 5GS, the NW-TT regenerates the Announce messages based on the Announce messages received from Follower port in NW-TT or DS-TT for the Leader ports in NW-TT and DS-TT(s). The NW-TT/UPF forwards the regenerated Announce messages to the PDU session(s) related to the Leader ports in the DS-TT(s).
NOTE 1: The TSN AF or TSCTSF can use the portDS.portState in the "Time synchronization information for each DS-TT port" element in UMIC to read and get notified for the port state changes for the PTP ports in DS-TT(s) and the portDS.portState in PMIC to read and get notified for the port state changes for the PTP ports in NW-TT. Based on the change of the port states, TSN AF or TSCTSF can determine that an external Grandmaster PTP Instance is found to be used instead of the GM in 5GS, or the GM in 5GS is selected as the Grandmaster PTP Instance and TSN AF or TSCTSF can disable or enable the (g)PTP grandmaster functionality in DS-TT(s), respectively.
NOTE 2: The TSN AF or TSCTSF can use the portDS.portState in PMIC to read and get notified for the port state changes for the PTP ports in DS-TT for the port state Disabled, Initializing or Faulty (if applicable). The TSCTSF or TSN AF can use the portDS.portEnable to indicate to the NW-TT that the DS-TT port is disabled. This avoids unnecessary (g)PTP traffic over User Plane to a DS-TT port in Disabled, Initializing or Faulty state. |
fbecc7f0dcf9784c6066646052ab0c0e | 23.501 | 5.27.1.7 Support for PTP grandmaster function in 5GS | The 5GS that is configured to operate as a time-aware system or Boundary Clock may support acting as a PTP grandmaster for a (g)PTP domain.
The configuration of PTP instances in DS-TT and NW-TT for PTP grandmaster function is described in clause K.2.
The following options may be supported (per DS-TT) for the 5GS to generate the Sync, Follow_Up and Announce messages for the Leader ports on the DS-TT:
a) NW-TT generates the Sync, Follow_Up and Announce messages on behalf of DS-TT (e.g. if DS-TT does not support this).
The NW-TT/UPF forwards the generated Sync, Follow_Up and Announce messages to the PDU session(s) related to the Leader ports on the DS-TT(s). The NW-TT timestamps the (g)PTP event message when the event message is sent to the PDU Session and adds TSi corresponding to the timestamp to the Sync message and the OriginTimestamp corresponding to the timestamp to Sync message (if one-step operation is used) or PreciseOriginTimestamp corresponding to the timestamp to Follow_Up message (if two-step operation is used) and sets the cumulative rateRatio value with 1. The OriginTimestamp or PreciseOriginTimestamp shall be set by NW-TT/UPF to the 5GS internal clock.
When DS-TT(s) receive the Sync, Follow_Up messages, it modifies the payload of the Sync, Follow_Up message as described for the PTP port in the egress TT in clause 5.27.1.2.2.2.
b) DS-TT generates the Sync, Follow_Up and Announce messages in this DS-TT. The OriginTimestamp or PreciseOriginTimestamp shall be set by DS-TT to the 5GS internal clock.
In both options, the NW-TT generates the Sync, Follow_Up and Announce messages for the Leader ports on the NW-TT. |
fbecc7f0dcf9784c6066646052ab0c0e | 23.501 | 5.27.1.8 Exposure of Time Synchronization | 5G System supports time synchronization service that can be activated and deactivated by AF. Exposure of time synchronization comprises the following capabilities:
- The AF may learn 5GS and/or UE availability and capabilities for time synchronization service.
- The AF controls activation and deactivation of the time synchronization service for the target UE(s).
- The AF may subscribe to time synchronization service status for the target UE(s).
The AF may use the service-specific parameters to control the time synchronization service for targeted UE(s). These parameters are specified in clause 4.15.9.3 and 4.15.9.4 of TS 23.502 [3] for (g)PTP-based and 5G access stratum-based time synchronization services, respectively.
The AF may subscribe for 5GS and/or UE availability and capabilities for time synchronization service. The AF indicates in the request the DNN, S-NSSAI and in addition the AF may indicate a list of UE identities or group identity to limit the subscription only to corresponding UEs. If the AF does not indicate DNN, S-NSSAI, the NEF determines the DNN, S-NSSAI based on the AF Identifier.
The TSCTSF (directly or via NEF) exposes the 5GS and/or UE availability and capabilities for synchronization service to the AF as described in clause 4.15.9.2 of TS 23.502 [3]. The exposed information includes the list of user plane node identities, the list of UE identities and may include the supported capabilities for (g)PTP time synchronization service per user plane node and UE.
The AF request to control the (g)PTP time synchronization service is sent to the TSCTSF (directly or via NEF). The request is targeted to a set of AF-sessions that are associated with the exposure of UE availability and capabilities for synchronization service.
The AF may request to use a specific PTP instance type when requesting the (g)PTP-based time synchronization distribution method (IEEE Std 1588 [126] or IEEE Std 802.1AS [104] operation (i.e. as a Boundary Clock, peer-to-peer Transparent Clock, or end-to-end Transparent Clock or as a PTP relay instance)). The request to control the (g)PTP time synchronization service may contain other service parameters as specified in Table 4.15.9.3-1 in clause 4.15.9.3 of TS 23.502 [3].
The AF may request to use the 5G access stratum as a time synchronization distribution method. In this case, the time source is provided by the 5GS. 5G-AN provides the 5GS time to the UE via 3GPP radio access; UE/DS-TT may provide 5G access stratum timing information to end stations using implementation specific means. The request to control the 5G access stratum time distribution (including the parameters such AF requests may contain) is described in clause 4.15.9.4 of TS 23.502 [3].
The AF or NEF selects the TSCTSF as specified in clause 6.3.24.
The AF request may include a time synchronization error budget (see also clause 5.27.1.9). The time synchronization error budget defines an upper bound for time synchronization errors introduced by 5GS.
The AF uses the procedure for configuring the (g)PTP instance in 5GS as described in clause 4.15.9.3 of TS 23.502 [3] and uses the procedure for providing the 5G access stratum time distribution as described in clause 4.15.9.4 of TS 23.502 [3] for the UEs.
The TSCTSF uses the Time Synchronization parameters (Table 4.15.9.3-1 in TS 23.502 [3]) as received from the AF (directly or via NEF) to control the (g)PTP time synchronization service. When IEEE Std 1588 [126] or IEEE Std 802.1AS [104] operation have been selected, the TSCTSF determines the necessary (g)PTP parameters to activate and control the service in DS-TT(s) and NW-TTs. For this purpose, the TSCTSF uses the PMIC or UMIC to manage the IEEE Std 1588 [126] or IEEE Std 802.1AS [104] operation in the DS-TT(s) or NW-TTs, respectively (see clause 5.27.1.4).
The TSCTSF may indicate whether it can support the service or not as per the requested acceptance criteria (e.g. based on the known timing synchronization status attribute thresholds also pre-configured at gNB) and provide notification when there is a service status update if the AF subscribes to service status updates (see also clause 5.27.1.12).
The TSCTSF uses the Time Synchronization parameters (Table 4.15.9.4-1 of TS 23.502 [3]) as received from the AF (directly or via NEF) to control the 5G access stratum time synchronization distribution as described in clause 4.15.9.4 of TS 23.502 [3].
For handling (g)PTP traffic, the PCF, according to PCC rule authorization, chooses a 5QI and dynamically set the PDB and/or MDBV according to requirements for (g)PTP protocol. The PCF provides the SMF with a PCC rule generated based on the AF request to control the (g)PTP time synchronization service. The SMF may take the information in the PCC rule to modify a PDU Session to create or modify or release a QoS Flow for transmitting the (g)PTP messages. The PCF acknowledges the policy request to the TSCTSF. The TSCTSF may report the result of the time synchronization request to the AF (directly or via NEF).
The AF may provide a temporal validity condition to the TSCTSF (directly or via NEF) when the AF activates or modifies the time synchronization service. Temporal validity condition contains the start-time and stop-time (in absolute time value) attributes that describe the time period when the time synchronization service is active for the targeted AF sessions. The TSCTSF manages the temporal validity condition as described in clauses 4.15.9.3 and 4.15.9.4 of TS 23.502 [3].
The AF may provide clock quality detail level and clock quality acceptance criteria to the TSCTSF (directly or via NEF) when the AF activates or modifies the time synchronization service. For ASTI based time synchronization services, the TSCTSF provides the clock quality reporting control information to AMF (see also clause 5.27.1.12).
The AF may provide a requested coverage area for the time synchronization service to the TSCTSF (directly or via NEF) when the AF activates or modifies the time synchronization service. The requested coverage area defines a spatial validity condition for the service using a geographical area (e.g. a civic address or shapes), or a list of Tracking Area Identities (TAIs) (see also clause 5.27.1.10). |
fbecc7f0dcf9784c6066646052ab0c0e | 23.501 | 5.27.1.9 Support for derivation of Uu time synchronization error budget | The AF may request a specific time synchronization error budget when requesting a time synchronization service employing the (g)PTP-based or 5G access stratum-based time distribution method. If the AF includes a time synchronization error budget in its request, the TSCTSF uses it to derive an error budget available for the NG-RAN to provide the 5G access stratum time via the Uu interface to each targeted UE (referred to as Uu time synchronization error budget hereafter).
The Time Synchronization Subscription data may optionally contain the authorized Uu time synchronization error budget. When the TSCTSF receives an AF request with a specific time synchronization error budget for the time synchronization service, the TSCTSF validates the Uu time synchronization error budget as described in clause 5.27.1.11.
To derive the Uu time synchronization error budget for each targeted UE, the TSCTSF takes the following into account:
- selected time synchronization distribution method (i.e. 5G access stratum-based time distribution or (g)PTP-based time distribution);
- Uu time synchronization error budget in the Time Synchronization Subscription data defined in clause 5.27.1.11;
- in the case of the (g)PTP-based time distribution:
- whether 5GS operates as a boundary clock and acts as a GM;
- whether a clock connected to the DS-TT/NW-TT acts as a GM;
- PTP port states;
- a CN part and a Device part of the time synchronization error budget (both parts may be predefined at the TSCTSF, or calculated by the 5GS using the implementation-specific means).
If the AF does not include a time synchronization error budget, the TSCTSF uses a preconfigured time synchronization error budget to derive the Uu time synchronization error budget. TheTSCTSF provides a 5G access stratum time distribution indication and the derived Uu time synchronization error budget to NG-RAN as described in clause 4.15.9.4 of TS 23.502 [3]. Based on this, NG-RAN provides the 5G access stratum time to the UE according to the Uu interface time synchronization error budget as provided by the TSCTSF (if supported by UE and NG-RAN). During Handover, Service Request, mobility registration and AM policy modification procedure, the AMF may provide the 5G access stratum time distribution indication and the Uu time synchronization error budget to NG-RAN as described in clause 4.15.9.4 of TS 23.502 [3], if needed.
NOTE: This release of the specification assumes that deployments ensure that the targeted UEs and the NG-RAN nodes serving those UEs support Rel-17 propagation delay compensation as defined in TS 38.300 [27]. |
fbecc7f0dcf9784c6066646052ab0c0e | 23.501 | 5.27.1.10 Support for coverage area filters for time synchronization service | This feature enables the AF to request time synchronization service for a UE or a group of UEs in a specific geographical area (so called coverage area). The requested coverage area contains a list of Tracking Area (TA) or a geographical area (e.g. a civic address or shapes that the NEF transforms to list of TAs based on pre-configuration).
TSCTSF checks with the UDM if the UE is allowed to receive the time synchronization service requested by AF and may update the Time Synchronization Coverage Area as described in clause 5.27.1.11.
The coverage area defines a spatial validity condition for the targeted UE(s) that is resolved at the TSCTSF. In order to do that, the TSCTSF may either:
a) discover the AMF(s) serving the list of TA(s) that comprise the spatial validity using Nnrf_NFDiscovery_Request service from the NRF and subsequently, the TSCTSF subscribes to the discovered AMF(s) to receive notifications about presence of the UE in an Area of Interest events (as described in clause 5.3.4.4). The subscription is targeted to Any UE. To reduce the number of Area of Interest reports (based on presence of UE) for the discovered AMFs, the TSCTSF may provide additional filtering information (e.g. List of UE IDs, DNN(s)/S-NNSAI(s)) to limit the subscription to the indicated UE identities, or UEs having a PDU Session with the given DNN/S-NSSAI as specified in clause 5.3.4.4.
b) determine the serving AMF for each of the targeted UE(s) using the UDM and subscribe to the serving AMF to receive notification about presence of the UE in an Area of Interest (as described in clause 5.3.4.4).
An Area of Interest (AoI) for each AMF is represented by a list of TA(s), wherein the Area of Interest is identical to the Time Synchronization Coverage Area.
Based on the outcome provided by the AMF about the UE's presence in the AoI, the TSCTSF determines if the time synchronization service is activated or deactivated.
For access stratum distribution activation/deactivation, the TSCTSF will enable/disable access stratum time distribution to the UE at the serving NG-RAN node reusing the procedures in clause 4.15.9.4 of TS 23.502 [3]. For (g)PTP distribution activation/deactivation, the TSCTSF will modify the PTP instance configuration by means of sending a PMIC to the impacted UE/DS-TTs and UMIC to the impacted UPF/NW-TT, as described in clause K.2.2.
If the time synchronization service is modified based on the reports of UE presence in the Area of Interest, the TSCTSF informs the AF for the impacted UE(s) by indicating the PTP port state for the related DS-TT PTP port (in the case of (g)PTP based time distribution) or notifying the AF with the indication of 5G access status time distribution status (enabled or disabled, for ASTI based time distribution).
If the Time Synchronization Coverage Area requested by the AF includes at least one TA that is a part of UE's Registration Area (RA), the 5GS may provide the AF-requested time synchronization service to the targeted UE within its RA, i.e. all TAs of the RA shall be treated as a Time Synchronization Coverage Area even if some of the TAs were not requested by the AF.
NOTE 1: This ensures that (1) there is no impact on the Registration Area (RA) of UE if/when the AMF receives a time synchronization service request with a spatial validity condition (i.e. for specific geographical area); and (2) the UE can continue receiving the time synchronization service when it moves within the RA in the RRC_IDLE state.
NOTE 2: Since the RA can be specific to the UE, the result can be different Time Synchronization Coverage Area for different DS-TT ports of different UEs within a PTP instance. |
fbecc7f0dcf9784c6066646052ab0c0e | 23.501 | 5.27.1.11 Controlling time synchronization service based on the Subscription | The distribution of timing information, 5G access stratum-based time distribution and (g)PTP-based time distribution, for a UE may be controlled based on subscription data stored in the UDM. The (g)PTP-based or 5G access stratum-based time synchronization service may be provided to a UE based on the UE's subscription which is specified in the TS 23.502 [3] clause 5.2.3.3.1.
The Access and Mobility Subscription data include the following information for the control of 5G access stratum-based time distribution:
- the Access Stratum Time Synchronization Service Authorization, which indicates whether the UE should be provisioned with 5G system internal clock timing information over access stratum as specified in TS 38.331 [28].
- optionally, the Uu time synchronization error budget.
- optionally, one or more periods of start and stop times defining the times when the UE should be provisioned with 5G system internal clock timing information.
- optionally, a Time Synchronization Coverage Area comprising a list of TAs where the UE shall be provisioned with 5G system internal clock timing information.
- optionally, a clock quality detail level indicating whether and which clock quality information to provide to the UE. It comprises one of the following values: clock quality metrics or acceptable/not acceptable indication.
- optionally, the clock quality acceptance criteria for the UE. It may be defined based on one or more attributes listed in Table 5.27.1.12-1.
During the Registration procedure, the AMF retrieves the subscription from UDM. If the AMF receives 5G access stratum-based time synchronization service subscription for the given UE, the AMF controls the 5G access stratum-based time distribution:
- If the 5G access stratum-based time synchronization service is allowed for the UE, the AMF provides the 5G access stratum time distribution indication to the NG-RAN so that it can provide 5G timing information to the UE.
- The AMF may provide a Uu time synchronization error budget to the NG-RAN (as described in clause 5.27.1.9). If the UE's subscription contains a Uu time synchronization error budget, then AMF sends it to NG-RAN. Otherwise, the AMF uses the pre-configured Uu time synchronization error budget and sends it to NG-RAN.
- If the UE's subscription contains Coverage Area (defined as a list of TAs), the AMF configures the NG-RAN to provide the 5G timing information to UE only when the UE is in the Coverage Area as described in clause 5.27.1.10.
- If the AMF receives the start and stop times, then the AMF enables and disables the 5G access stratum time distribution indication to the NG-RAN according to the expiry of start and stop times if the UE is in CM-CONNECTED state. If the UE is in CM-IDLE state when a Start time condition is met, the AMF pages the UE and provides the 5G access stratum time distribution indication to NG-RAN as part of the subsequent service request procedure initiated by the UE in the response to the paging.
- If the AMF receives the clock quality detail level, then the AMF configures the NG-RAN to provide clock quality detail information reporting to UE as described in clause 5.27.1.12. The AMF may instruct the UE to reconnect to the network when the UE detects that the RAN timing synchronization status has changed while the UE is in RRC_INACTIVE or RRC_IDLE, as described in clause 5.27.1.12.
- If the AMF receives the same parameters both in the Access and Mobility Subscription data from UDM and in the AM Policy from PCF, the AMF shall use the value received from the AM policy.
The Time Synchronization Subscription data is the subscription data for the control of (g)PTP-based time distribution and 5G access stratum-based time distribution and includes the following information:
- the "AF request Authorization", indicating whether the UE is authorized for an AF-requested 5G access stratum-based time distribution and (g)PTP-based time distribution services. The indication is provided separately for each service:
- "allowed" or "not allowed" for (g)PTP based time synchronization service (per DNN/S-NSSAI and UE identity),
- "allowed" or "not allowed" for ASTI based time synchronization services (per UE identity).
- If the "AF request Authorization" is set to "allowed", the Time Synchronization Subscription data may contain additional information for authorization of (g)PTP- and ASTI- based time synchronization services: this information is optional and may be provided separately for authorization of each service:
- optionally, a list of TA(s) which specifies an area (a so-called Authorized Time Synchronization Coverage Area) in which an AF may request time synchronization services;
- optionally, one or more periods of authorized start and stop times, which indicates the allowed time period during which an AF may request time synchronization services;
- optionally, authorized Uu time synchronization error budget, which indicates the limit the AF may request.
- optionally, "allowed" or "not allowed" for clock quality detail level equals to "clock quality metrics", which indicates whether clock quality metrics information may be provided to the UE.
- optionally, "allowed" or "not allowed" for clock quality detail level equals to "acceptable/not acceptable indication", which indicates whether an acceptable/not acceptable indication may be provided to the UE.
- optionally, one or more sets of clock quality acceptance criteria for the UE using the TSS attributes as defined in clause 5.27.1.12.
NOTE 1: If the clock quality detail level set to "acceptable/not acceptable indication" is "allowed", clock quality acceptance criteria need to be provided.
- one or more Subscribed time synchronization service ID(s), each containing the DNN/S-NSSAI and a reference to a PTP instance configuration pre-configured at the TSCTSF (e.g. PTP profile, PTP domain, etc.):
- optionally, for each PTP instance configuration, one or more periods of start and stop times defining active times of time synchronization service for the PTP instance.
- optionally, for each PTP instance configuration, a Time Synchronization Coverage Area defining a list of TAs where the (g)PTP-based time synchronization is available for the UEs in the PTP instance.
- optionally, for each PTP instance configuration, Uu time synchronization error budget.
The TSCTSF retrieves the Time Synchronization Subscription data from UDM. If the TSCTSF receives the Time Synchronization Subscription data for a UE, the TSCTSF controls the Time Synchronization Service including (g)PTP-based time distribution and 5G access stratum-based time distribution:
- The TSCTSF retrieves the Time Synchronization Subscription data from the UDM when the TSCTSF receives an AF request for the time synchronization service (either ASTI or (g)PTP). According to the "AF request Authorization" in the UE's Time Synchronization Subscription data, the TSCTSF determines whether the UE is authorized for an AF-requested time synchronization service:
- If the UE's Time Synchronization Subscription data contains an Authorized Time Synchronization Coverage Area (i.e. a list of TA(s) defining the restricted area for AF request), TSCTSF checks whether the AF requested Coverage Area satisfies the authorized area: If the requested Coverage Area (see clause 5.27.1.10) is within the Authorized Time Synchronization Coverage Area, the TSCTSF uses the requested Coverage Area. If the Authorized Time Synchronization Coverage Area is inside of the requested Coverage Area, the TSCTSF uses the Authorized Time Synchronization Coverage Area. If the requested Coverage Area partly overlaps with the Authorized Time Synchronization Coverage Area, the TSCTSF uses the intersection of them. If there is no overlap between them, the TSCTSF shall reject the AF request.
- If the AF requested Coverage Area satisfies the authorized area totally or partly, TSCTSF notifies to AF with the Time Synchronization Service information based on the Authorized Time Synchronization Coverage Area. TSCTSF subscribes to UE's presence in the Area of Interest at the discovered AMF(s), if the UE(s) moves out of the Time Synchronization Coverage Area requested by AF or updated by TSCTSF, the TSCTSF shall disable Time Synchronization Service and notifies to AF.
- If the UE's Time Synchronization Subscription data contains authorized Uu time synchronization error budget, the TSCTSF checks whether the Uu time synchronization error budget derived from AF request satisfies (i.e. equal or larger than) the authorized Uu time synchronization error budget.
- If the UE's Time Synchronization Subscription data contains periods of authorized start and stop times, the TSCTSF checks whether the AF requested temporal validity condition satisfies (i.e. within) any of the periods of authorized start and stop times. If such period is found, the TSCTSF uses the start and stop times of the AF request.
- If the UE's Time Synchronization Subscription data contains a clock quality detail level (and clock quality acceptance criteria, if applicable), the TSCTSF checks whether the AF-requested clock quality information to be reported complies with the authorized clock quality detail level and the clock quality acceptance criteria (if applicable) in the UE's Time Synchronization Subscription data.
Each parameter in the AF-requested clock quality is evaluated individually, if the TSCTSF determines that at least one parameter is "not acceptable", the TSCTSF rejects the AF request.
If the "Traceable to GNSS" in the subscription data is "Yes", the AF request for Traceable to GNSS is allowed. Otherwise, the AF request for Traceable to GNSS is not allowed.
If the "Traceable to UTC" in the subscription data is "Yes", the AF request for Traceable to UTC is allowed. Otherwise, the AF request for Traceable to UTC is not allowed.
For Frequency stability and Clock Accuracy, AF is not allowed to request value which is lower than value in subscription data.
- If the AF request is authorized, the TSCTSF proceeds as specified in clause 5.27.1.8 and in TS 23.502 [3]. Otherwise, the TSCTSF rejects the AF request.
- The TSCTSF retrieves the Time Synchronization Subscription data from the UDM when it receives notification from the PCF that a UE has established a PDU Session that is potentially impacted by (g)PTP-based time synchronization service:
- The TSCTSF retrieves the PTP instance configurations referenced from the "Subscribed time synchronization service ID(s)". The PTP instance configurations are stored locally in the TSCTSF. The TSCTSF determines if one or more of the PTP instance configurations match with the DNN/S-NSSAI of the given PDU Session. If no PTP instance exists for the given PTP instance configuration, the TSCTSF initializes the PTP instance in 5GS as described in clause K.2.2.
- The TSCTSF configures a PTP port in DS-TT and adds it to the corresponding PTP instance in NW-TT as described in clause K.2.2.
- If the PTP instance configuration referenced by UE's Time Synchronization Subscription data contains an Uu time synchronization error budget, then the TSCTSF uses it to derive an Uu time synchronization error budget available for the NG-RAN to provide the 5G access stratum time for the UE as specified in clause 5.27.1.9.
- If the PTP instance configuration referenced by the Time Synchronization Subscription data for the UE contains start and stop times, the TSCTSF, upon expiry of start time, creates the PTP instance and adds the PTP port in DS-TT to the PTP instance. Upon expiry of stop time, if this is the last period of start and stop times in the PTP instance configuration, the TSCTSF deletes the PTP instance, otherwise the TSCTSF temporarily removes the PTP port in DS-TT from the corresponding PTP instance as specified in clause 4.15.9.3 of TS 23.502 [3].
- If the PTP instance configuration referenced by the Time Synchronization Subscription data for the UE contains a Time Synchronization Coverage Area, the TSCTSF subscribes to UE's Presence in Area(s) of Interest corresponding to the Time Synchronization Coverage Area at the discovered AMF(s). When the TSCTSF determines that the UE has moved inside or outside of the Time Synchronization Coverage Area, the TSCTSF adds or temporarily removes the PTP port in DS-TT from the corresponding PTP instance. |
fbecc7f0dcf9784c6066646052ab0c0e | 23.501 | 5.27.1.12 Support for network timing synchronization status monitoring | While the time synchronization service is offered by the 5GS, based on 5G access stratum-based time distribution or (g)PTP-based time distribution, the network timing synchronization status of the nodes involved in the operation (e.g. gNBs and/or UPF/NW-TTs) may change. gNBs and UPF/NW-TT can detect timing synchronization degradation or improvement locally. The support for network timing synchronization status monitoring enables the 5GS to modify time synchronization service for a UE or a group of UEs depending on the current synchronization status and notify service updates. There may be three consumers of this information:
- TSCTSF may receive node-level information about timing synchronization status from gNB and/or UPF/NW-TT directly from OAM or alternatively, if supported by a node, using control plane signalling at node level. Node level signalling uses UMIC for UPF/NW-TT case and an AMF service to report N2 node level information for the gNB case. In the latter case, the TSCTSF may provide a list of gNB IDs or a list of TAs in the subscription request for RAN timing synchronization status reporting and the AMF controls the gNB node level reporting and subscription using NGAP messages (see TS 38.413 [34]).
- AF may subscribe to time synchronization status notifications for a UE or group of UEs for which the AF requests or has requested time synchronization service (for 5G access stratum time distribution or (g)PTP services).
- For 5G access stratum time synchronization service, the UE may receive clock quality information from the gNB based on UE subscription data stored in the UDM (see clause 5.27.1.11) or AF request for clock quality reporting to the UE.
When activating time synchronization for a UE, TSCTSF forwards the clock quality detail level (if available) to the AMF (via PCF using AM Policy Association Modification).
If the UE indicates its support for reconnection to the network due to changes in RAN timing synchronization status as specified in clause 5.4.4a and the AMF receives Clock Quality Reporting Control Information (CQRCI) either from TSCTSF via PCF or from the UDM as part of the subscription data, the AMF instructs the UE to transition to the RRC_CONNECTED state when the UE detects that the gNB timing synchronization status has changed while the UE is in the RRC_INACTIVE or RRC_IDLE state. When the UE wants to access the 5GS, the UE shall perform Unified Access Control as defined in TS 38.331 [28].
gNBs and TSCTSF may be pre-configured with thresholds for each Timing Synchronization Status (TSS) attribute, if supported, that is described in Table 5.27.1.12-1.
gNBs may include a reference report ID in SIB information, if supported. A reference report ID consists of a scope of the TSS and an Event ID. A scope of the TSS supports providing TSS information for all cells or a group of cells within a single gNB. Event ID is an integer indicating that the gNB's clock quality has changed, resulting in at least one TSS attribute exceeding or meeting again the pre-configured threshold. Uniqueness of Event ID value is ensured by combining it with a gNB ID as specified in TS 38.300 [27].
The RAN timing synchronization status information includes the gNB node-level information about timing synchronization operation status, as described in TS 38.300 [27]. When the gNB is monitoring RAN timing synchronization status, the gNB determines the reporting triggers as described in TS 38.401 [42].
The gNB notifies the TSCTSF (either using N2 node level signalling via AMF, or via OAM) about timing synchronization status changes based on the pre-configured thresholds with the scope of the timing synchronization status (i.e. gNB ID or a list of Cell IDs within a single gNB) and the corresponding network timing synchronization status attributes as described in Table 5.27.1.12-1. The gNB may support only one or more network timing synchronization status attributes. Upon reception of the notification, the TSCTSF determines whether the TSS attribute meets (status improvement) or exceeds (status degradation) the preconfigured threshold value.
The gNB indicates the status change to the UEs via the reference report ID change in SIB information:
- When the network timing synchronization status exceeds any of the pre-configured thresholds or meets the threshold again, the gNB changes the reference report ID in SIB information. Either event serves as a notification for the UEs reading the SIB information that there is new TSS information available.
NOTE 1: NG-RAN is assumed not to provide clock quality metrics better than the pre-configured threshold, i.e. if a clock quality metric is better than the corresponding threshold, the NG-RAN reports the threshold value to the UE in an RRC message instead.
NOTE 2: It is assumed the pre-configured thresholds in the gNB(s) are sufficient to meet UE time sync performance requirement which are configured by the operator.
- If supported, the UE in the RRC_INACTIVE or RRC_IDLE state compares the reference report ID in SIB information with its locally stored reference report ID to determine whether it has the latest available clock quality information already or it needs to transit to the RRC_CONNECTED state to retrieve it.
- If the UE is instructed by AMF (via the Registration procedure, or the UE Configuration Update procedure) to reconnect to the network in the case when the UE determines that the reference report ID has changed, the UE in the RRC_INACTIVE or RRC_IDLE state, if supported by the UE, reconnects to the network. RAN may delay or prioritize UE's transition to the RRC_CONNECTED state using the UAC framework [28], i.e. UEs are not expected to transition to the RRC_CONNECTED state immediately after determining that the clock quality information has changed and receiving instructions from the AMF. After the UE has reconnected to the network, the gNB uses unicast RRC signalling to provision the clock quality information to the UEs.
The network timing synchronization status information from gNB or UPF/NW-TT to the TSCTSF may contain the following information as described in the Table 5.27.1.12-1. The details for gNB timing synchronization status information are specified in TS 38.413 [34]. However, it is up to gNB to determine whether to provide its timing synchronization status reporting and which of the information elements to include in the TSS report to the TSCTSF, i.e. based on the implementation gNB may report all or some, or none of the information elements from Table 5.27.1.12-1.
Table 5.27.1.12-1: Information elements that gNB or UPF/NW-TT timing synchronization status information may contain (all optional)
Information Name
Description
Synchronization state
Indicates the state of the node synchronization, represented by the values "Locked", "Holdover", or "Freerun" (NOTE 1).
Clock quality
>> Traceable to GNSS
Indicates whether the current time source is traceable to the GNSS and represented by values “Yes” or “No”.
>> Traceable to UTC
Indicates whether the current time source is traceable to the UTC and represented by values “Yes” or “No”.
>> Frequency stability
Describes the estimate of the variation of the local clock when it is not synchronized to another source (NOTE 2).
>> Clock Accuracy
Describes the mean in ns over an ensemble of measurements of the time between the clock under test and a reference clock (NOTE 3).
Parent time source
Describes the primary source the node is currently using, represented by the values "SyncE", "PTP", "GNSS", "atomic clock", "terrestrial radio", "serial time code", "NTP", "hand_set", "other".
NOTE 1: Clock is in the "Locked", "Holdover", or "Freerun" mode, as defined in ITU‑T G.810 [164].
NOTE 2: Frequency stability is estimated in a similar manner as for offsetScaledLogVariance attribute defined in clause 7.6.3.5 of IEEE Std 1588 [126].
NOTE 3: Clock accuracy measurement considers accuracy up to gNB antenna and RAN internal process.
The TSCTSF determines the UEs impacted by gNB's timing synchronization status change or UPF timing synchronization status change (only for the case when UPF/NW-TT is involved in providing time information to DS-TT) by comparing the received timing synchronization status information with a clock quality acceptance criteria provided by the AF. The TSCTSF evaluates whether the received timing synchronization status is acceptable or not acceptable for a UE.
Each TSS attribute in the clock quality acceptance criteria is evaluated individually, if the TSCTSF determines that at least one reported value is "not acceptable" or there is at least one attribute present in the clock quality acceptance criteria is not reported, then the TSCTSF shall consider the timing synchronization status as "not acceptable" for the UE.
- For the gNB case, when the TSCTSF receives information about timing synchronization status change, the TSCTSF uses the NRF to discover the AMFs serving the impacted gNBs and subscribes to receive notifications for UE's presence in Area of Interest information from AMF as described in clause 5.3.4.4. The Area of Interest is set to the scope of the timing synchronization status (i.e. gNB ID or a group of cells within the gNB specified with a list of Cell IDs that has reported status degradation (i.e. the pre-configured thresholds are exceeded in the gNB). The subscription is targeted to any UE in the AMF, the TSCTSF may provide additional filtering information as specified in clause 5.3.4.4 (e.g. List of UE IDs, DNN(s)/S-NNSAI(s)) to limit the subscription to the indicated UE identities, UEs having a PDU Session with the given DNN(s)/S-NSSAI(s). The TSCTSF correlates information about impacted gNBs and the UE location information received from the AMF. If the gNB notifies the TSCTSF for the status improvement (i.e. the pre-configured thresholds are met in the gNB), the TSCTSF modifies the subscription to remove the corresponding Area of Interest from the subscription.
- For UPF case, the TSCTSF determines the UEs for which the impacted UPF/NW-TT is configured to send (g)PTP messages on behalf of DS-TT (see clause 5.27.1.7).
If the gNB's or UPF's timing synchronization status change, the TSCTSF may perform the following:
- For AFs that subscribe for 5G access stratum time synchronization service or (g)PTP time synchronization service status update (i.e. change in support status of the clock quality acceptance criteria provided by the AF and specified using TSS attributes from Table 5.27.1.12-1), the TSCTSF may provide notification towards the AF when there is a change in support status for a UE or group of UEs.
- Deactivating/reactivating/updating time synchronization services:
- (g)PTP time synchronization service case: For UEs that are part of a PTP instance and which are impacted by NG-RAN or UPF time synchronization status degradation or improvement:
- If TSCTSF determines that the clock quality acceptance criteria provided by AF can still be met, then TSCTSF may update the clock quality information sent in Announce messages (see clause 7.6.2 of IEEE 1588 [8]) for the PTP instance using existing procedures and existing PMIC/UMIC information. The handling of Announce messages follows existing procedures as described in clause 5.27.1.6.
- If TSCTSF determines that the clock quality acceptance criteria provided by AF cannot be met, then TSCTSF informs the AF for the corresponding PTP ports being inactive due to the result of fulfilling the clock quality acceptance criteria; and the TSCTSF temporarily removes the UE/DS-TT from the PTP instance using the procedure in clause K.2.2.1 and clause K.2.2.4. The AF may send a service update or delete request (see clause 4.15.9.3 of TS 23.502 [3]).
- If TSCTSF determines that the clock quality acceptance criteria provided by AF can be met again then TSCTSF informs the AF about the result, adds the DS-TT PTP port to the PTP instance again and re-activates the Grandmaster functionality.
For 5G access stratum time synchronization service, clock quality reporting control information (CQRCI) manages the gNBs timing synchronization status reporting to the UE. The TSCTSF may retrieve CQRCI from UDM or receive CQRCI from AF request. The TSCTSF needs to check whether the AF requested CQRCI is allowed as defined in clause 5.27.1.11. When AMF provides the 5G access stratum time distribution indication and the Uu time synchronization error budget to gNB, the AMF also includes the clock quality reporting control information (CQRCI) provided by the TSCTSF via AM policy or retrieved from UDM. If the AMF receives CQRCI both in the Access and Mobility Subscription data from UDM and in the AM Policy from PCF, the AMF shall use the value received from the AM policy. CQRCI contains the following fields:
- Clock quality detail level. It indicates whether and which clock quality information to provide to the UE and can take one of the following values: "clock quality metrics" or "acceptable/not acceptable indication".
- If the clock quality detail level equals "clock quality metrics", the NG-RAN provides clock quality metrics to the UE that reflect its current timing synchronization status. i.e. one or more of the following information elements: clock accuracy, traceability to UTC, traceability to GNSS, frequency stability, parent time source, synchronization state as defined in Table 5.27.1.12-1. NG-RAN is locally configured which of the clock quality metrics supported by NG-RAN are provided to UE(s).
- If the clock quality detail level equals "acceptable/not acceptable indication", NG-RAN provides clock quality acceptance criteria for the UE. The gNB provides an acceptable indication to the UE if the gNB's timing synchronization status matches the acceptance criteria received from the AMF; otherwise, the gNB indicates "not acceptable" to the UE. Clock quality acceptance criteria can be defined based on one or more information elements listed in Table 5.27.1.12-1. If AF includes clock quality acceptance criteria in its request towards TSCTSF, the AF shall be notified about the result once TSCTSF determines whether the clock quality acceptance criteria can be met or not. Based on the notification, the AF may decide to modify the service if preferred (e.g. disable the service upon status degradation or enable it again upon status improvement).
When determining the clock quality metrics for a UE and when determining whether clock quality is acceptable or not acceptable for a UE, the gNB considers whether propagation delay compensation is performed.
NOTE 3: In this Release, UE capabilities and internal inaccuracies are assumed to be budgeted by the client network operator when agreeing the required clock accuracy with the 5G network operator.
To provision clock quality information to the UEs, a gNB uses unicast RRC signalling:
- For UEs in the RRC_CONNECTED state, the gNB uses unicast RRC signalling.
- UEs that are not in the RRC_CONNECTED state first need to establish or resume the RRC connection to receive the clock quality information from the gNB via unicast RRC signalling.
During N2 Handover and Xn handover, Service Request, mobility registration and AM policy modification procedure, the AMF may provide the CQRCI to NG-RAN.
5.27.1a Periodic deterministic communication
This clause describes 5G System features that allow support of periodic deterministic communication where the traffic characteristics are known a-priori and a schedule for transmission from the UE to a downstream node, or from the UPF to an upstream node is provided via external protocols outside the scope of 3GPP (e.g. IEEE 802.1 TSN).
The features include the following:
- Providing TSC Assistance Information (TSCAI) that describe TSC flow traffic characteristics (as described in clause 5.27.2) at the gNB ingress and the egress of the UE for traffic in downlink and uplink direction, respectively;
- Support for hold & forward buffering mechanism (see clause 5.27.4) in DS-TT and NW-TT to de-jitter flows that have traversed the 5G System. |
fbecc7f0dcf9784c6066646052ab0c0e | 23.501 | 5.27.2 TSC Assistance Information (TSCAI) and TSC Assistance Container (TSCAC) | |
fbecc7f0dcf9784c6066646052ab0c0e | 23.501 | 5.27.2.1 General | TSC Assistance Information (TSCAI) is defined in Table 5.27.2-1 and describes TSC traffic characteristics for use in the 5G System. It can also be used for services such as XR services (AR/VR applications) and interactive media services as specified in clause 5.37.
TSCAI may be used by the 5G-AN, if provided by SMF. The knowledge of TSC traffic pattern is useful for 5G-AN as it allows more efficiently scheduling of QoS Flows that have a periodic, deterministic traffic characteristics either via Configured Grants, Semi-Persistent Scheduling or with Dynamic Grants. TSCAI can be provided for both GBR and non-GBR QoS flows.
The TSCTSF determines the TSC Assistance Container (defined in Table 5.27.2-2) based on information provided by an AF/NEF or a DetNet controller as described in clause 5.27.2.3 and provides it to the PCF for IP type and Ethernet type PDU Sessions. In the case of integration with IEEE TSN network, the TSN AF determines TSC Assistance Container as described in clause 5.27.2.2 and provides it to the PCF for Ethernet PDU Sessions. The PCF receives the TSC Assistance Container from the TSCTSF or the TSN AF and forwards it to the SMF as part of PCC rule as described in clause 6.1.3.23a of TS 23.503 [45].
The SMF binds a PCC rule with a TSC Assistance Container to a QoS Flow as described in clause 6.1.3.2.4 of TS 23.503 [45]. The SMF uses the TSC Assistance Container to derive the TSCAI for that QoS Flow and sends the derived TSCAI to the NG-RAN. The Periodicity, Periodicity Range, Burst Arrival Time (BAT), BAT Window and Survival Time components of the TSCAI are specified by the SMF with respect to the 5G clock. The SMF is responsible for mapping the Burst Arrival Time, BAT Window, Periodicity and Periodicity Range from an external clock (when available) to the 5G clock based on the time offset and cumulative rateRatio (when available) between the external clock time and 5GS time as measured and reported by the UPF. The SMF determines the TSCAI as described in clause 5.27.2.4.
A Survival Time, which indicates the time period an application can survive without any data burst, may be provided by TSN AF/AF or by the TSCTSF either in terms of maximum number of messages (message is equivalent to all packets of a data burst) or in terms of time units. Only a single data burst is expected within a single time period referred to as the periodicity.
The SMF may send an update of the TSCAI to the NG-RAN as defined in clauses 4.3.3.2, 4.9.1.2.2 and 4.9.1.3.2 of TS 23.502 [3] or as defined in clause 5.37.8.2.
Table 5.27.2-1: TSC Assistance Information (TSCAI)
Assistance Information
Description
Flow Direction
The direction of the flow (uplink or downlink).
Periodicity
It refers to the time period between start of two data bursts.
Burst Arrival Time (optional)
The latest possible time when the first packet of the data burst arrives at either the ingress of the RAN (downlink flow direction) or the egress of the UE (uplink flow direction).
Survival Time (optional)
Survival Time, as defined in TS 22.261 [2], refers to the time period an application can survive without any data burst.
Burst Arrival Time Window (BAT Window) (optional)
(NOTE 1) (NOTE 2)
Indicates the acceptable earliest and latest arrival time of the first packet of the data burst at either the ingress of the RAN (downlink flow direction) or the egress of the UE (uplink flow direction).
Capability for BAT adaptation (optional) (NOTE 1)
Indicates that the AF will adjust the burst sending time according to the network provided Burst Arrival Time offset (see clause 5.27.2.5).
N6 Jitter Information (optional)
(NOTE 3)
Jitter information associated with the Periodicity in downlink (see clause 5.37.8.1).
Periodicity Range (optional) (NOTE 4)
It indicates that the AF will adjust the periodicity and provides the acceptable range (which is formulated as lower bound and upper bound of the Periodicity) or acceptable Periodicity value(s) (which is formulated as a list of values for the Periodicity).
NOTE 1: Only one of the parameters (BAT Window or Capability for BAT adaptation) can be provided.
NOTE 2: The parameter can only be provided together with Burst Arrival Time.
NOTE 3: Only one of the parameters Burst Arrival Time or N6 Jitter Information may be provided for a given Traffic Flow.
NOTE 4: The Periodicity Range can only be provided together with Periodicity when Burst Arrival Time and Burst Arrival Time Window are present.
Table 5.27.2-2: TSC Assistance Container (TSCAC)
Assistance Information
Description
Flow Direction
The direction of the TSC flow (uplink or downlink).
Periodicity
It refers to the time period between start of two data bursts.
Burst Arrival Time (optional)
The time when the first packet of the data burst arrives at the ingress port of 5GS for a given flow direction (DS-TT for uplink, NW-TT for downlink).
Survival Time (optional)
It refers to the time period an application can survive without any data burst, as defined in TS 22.261 [2].
Time Domain (optional)
The (g)PTP domain of the TSC flow.
Burst Arrival Time Window (BAT Window) (optional)
(NOTE 1) (NOTE 2)
Indicates the acceptable earliest and latest arrival time of the first packet of the data burst at the ingress port of 5GS for a given flow direction (DS-TT for uplink, NW-TT for downlink).
Capability for BAT adaptation (optional) (NOTE 1)
It indicates that the AF will adjust the burst sending time according to the network provided Burst Arrival Time offset (see clause 5.27.2.5).
Periodicity Range (optional) (NOTE 3)
It indicates that the AF will adjust the periodicity and provides the acceptable range (which is formulated as lower bound and upper bound of the Periodicity) or acceptable Periodicity value(s) (which is formulated as a list of values for the Periodicity).
NOTE 1: Only one of the parameters (BAT Window or Capability for BAT adaptation) can be provided.
NOTE 2: The parameter can only be provided together with Burst Arrival Time.
NOTE 3: The Periodicity Range can only be provided together with Periodicity when Burst Arrival Time and Burst Arrival Time Window are present. |
fbecc7f0dcf9784c6066646052ab0c0e | 23.501 | 5.27.2.2 TSC Assistance Container determination based on PSFP | In the case of integration with IEEE TSN network, the TSN AF determines a TSC Assistance Container (defined in Table 5.27.2-2) and provides it to the PCF. The determination of TSC Assistance Container based on Per-Stream Filtering and Policing (PSFP) information applies only to Ethernet type PDU Sessions.
NOTE 1: This clause assumes that PSFP information as defined in IEEE Std 802.1Q [98] and Table K.3.1-1is provided by CNC. PSFP information may be provided by CNC if TSN AF has declared PSFP support to CNC. TSN AF indicates the support for PSFP to CNC only if all the DS-TT and NW-TT ports of the 5GS Bridge have indicated support of PSFP. Means to derive the TSC Assistance Container if PSFP is not supported by 5GS and/or the CNC are beyond the scope of this specification.
The TSN AF may be able to identify the ingress port and thereby the PDU Session as described in clause 5.28.2.
The TSN AF interfaces towards the CNC for the PSFP (IEEE Std 802.1Q [98]) managed objects that correspond to the PSFP functionality implemented by the DS-TT and the NW-TT. Thus, when PSFP information is provided by the CNC, the TSN AF may extract relevant parameters from the PSFP configuration. The TSN AF calculates traffic pattern parameters (such as burst arrival time with reference to the ingress port and periodicity). TSN AF also obtains the flow direction as specified in clause 5.28.2. Survival Time may be pre-configured in TSN AF.
TSN AF may enable aggregation of TSN streams if the TSN streams belong to the same traffic class, terminate in the same egress port and have the same periodicity and compatible Burst arrival time. When Survival Time information is provided for a TSN stream, then it should not be aggregated with other TSN streams into a single QoS Flow, or if they are aggregated, then the Survival Time parameter shall not be provided. One set of parameters and one TSC Assistance Container are created by the TSN AF for multiple TSN streams to enable aggregation of TSN streams to the same QoS Flow.
Annex I describe how the traffic pattern information is determined.
NOTE 2: Further details of aggregation of TSN streams (including determination of burst arrival times that are compatible so that TSN streams can be aggregated) are left for implementation.
NOTE 3: In order for the TSN AF to get Burst Arrival Time, Periodicity on a per TSN stream basis, support for IEEE Std 802.1Q [98] (as stated in clause 4.4.8.2) Per-Stream Filtering and Policing (PSFP) with stream gate operation is a prerequisite.
For a UE-UE TSC stream, the (TSN) AF divides the stream into one uplink stream and one or more downlink streams as defined in clause 5.28.2. The TSN AF binds the uplink and downlink streams to the PDU Sessions and provides the streams on AF Session basis to the PCF(s). The TSN AF calculates traffic pattern parameters for the UL and the DL stream using the PSFP configuration (if provided) respectively:
- For the uplink stream, the Flow Direction is set to uplink and traffic pattern parameters (such as burst arrival time with reference to the ingress port and periodicity) is determined as described in Annex I.
- For downlink stream, the Flow Direction is set to downlink, the burst arrival time is set to sum of burst arrival time of the UL stream and 5GS Bridge delay of PDU Session carrying the UL stream and the periodicity is determined as described in Annex I. |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.