4.2.3 Starting a trace recording session – signalling based

32.4223GPPRelease 17Subscriber and equipment traceTelecommunication managementTrace control and configuration managementTS

Tools: ARFCN - Frequency Conversion for 5G NR/LTE/UMTS/GSM

4.2.3.1 UTRAN starting mechanisms

In an RNC the Trace Recording Session will always be the same as the Trace Session as no triggering events are defined in UTRAN. Therefore a Trace Recording Session should be started in an SRNC when the SRNC receives the CN_INVOKE_TRACE message from the Core Network and if some activities have been started on the interfaces that have been requested to be traced.

The CN_INVOKE_TRACE message that is received from the Core Network (MSC Server or SGSN) contains the following information:

– Trace Reference

– Trace Target: UE identity (IMSI or IMEI(SV))

– Trace Recording Session Reference

– Trace Depth

– List of Interfaces for RNC

– Job Type

– Area Scope

– List of Measurements

– Reporting Trigger

– Report Interval

– Report Amount

– Event Threshold

– Logging Interval

– Logging Duration

– Trace Collection Entity IP Address

– MDT PLMN List

Note that at the same time not all of the parameters can be present. The conditions which parameters shall be present is described in clause 5 of the present document.

If the Job Type parameter indicates MDT (e.g. Immediate or logged MDT is required) in the CN_INVOKE_TRACE message the RNC shall also configure MDT to the UE. The detailed mechanism of the MDT configuration to the UE is defined in TS 37.320 [30] and TS 25.331 [31].

The RNC shall send the following parameters to the UE in case of Logged MDT:

– Trace Reference

– Trace Recording Session Reference

– Area Scope

– TCE Id (The value signalled as IP address of TCE is mapped to a TCE Id, using a configured mapping in the RNC).

– Logging Interval

– Logging Duration

– MDT PLMN List

In case of Immediate MDT the RNC shall send the following parameters to the UE:

– List of Measurements

– Reporting Trigger

– Report Interval

– Report Amount

– Event Threshold

Note that at the same time not all of the parameters can be present. The conditions which parameters shall be present is described in clause 5 of the present document.

If the SRNC does not have enough resources it may not start a Trace Recording Session.

The Trace Recording Session Reference shall be the same as received in the CN_INVOKE_TRACE message.

In a DRNC the Trace Recording Session should be started when the DRNC receives the IUR_INVOKE_TRACE message. If the DRNC does not have enough resources it may not start a Trace Recording Session.

The Trace Session is activated to the RNC by sending a CN_INVOKE_TRACE message from the CN (MSC Server or SGSN). When RNC receives the CN_INVOKE_TRACE message it should immediately start a Trace Session and a Trace Recording Session according to the trace control and configuration parameters received in the CN_INVOKE_TRACE message.

If there are not enough resources in RNC to start a Trace Recording Session, the RNC may reject to start a Trace Recording Session. However the RNC shall start the Trace Session. Session and if MDT activation is requested shall activate the MDT to the UE.

When SRNC/DRNC receives the trace control and configuration parameters:

– If the Trace Reference is the same as an existing Trace Session for the same subscriber or equipment in SRNC/DRNC, the SRNC/DRNC shall not activate a new Trace Session and the existing Trace Session will not be impacted;

– If the Trace Recording Session Reference is the same as an existing Trace Recording Session in the existing Trace Session having the same Trace Reference, the SRNC/DRNC shall not start a new Trace Recording Session;

– If the trace control and configuration parameters are received from the same CN domain (CS/PS) as the existing Trace Recording Session(s) if any, and the Trace Recording Session Reference is not the same as any existing Trace Recording Session(s) in the existing Trace Session having the same Trace Reference, the SRNC/DRNC shall start a new Trace Recording Session;

– If the trace control and configuration parameters are received from different CN domain (CS/PS) as the existing Trace Recording Session(s) if any (i.e. the RNC has simultaneoud CS and PS connection and CN_INVOKE_TRACE is received from both connection), regardless of the Trace Recording Session Reference is the same or not as any existing Trace Recording Session(s) in the existing Trace Session having the same Trace Reference, the SRNC shall not start a new Trace Recording Session;

– If the Trace Reference is the same as an existing Trace Session for different subscriber(s) or equipment(s) in SRNC/DRNC, the SRNC /DRNC shall not activate a new Trace Session, and the SRNC/DRNC shall not start a new Trace Recording Session.

The following figure shows an example for a CS call how the Trace Session is activated to RNC. In the example it is assumed that there is no PS connection at all during the CS call.

Figure 4.2.3.1.1: Starting a Trace Recording Session (Signalling) in UTRAN

Interaction with soft-handovers

If the subscriber or equipment, which is traced, makes a soft handover the SRNC should propagate the trace control and configuration parameters (including MDT specific parameters if they are available) further to the DRNC by using the IUR_INVOKE_TRACE message. When the DRNC receives the IUR_INVOKE_TRACE message it should immediately start a Trace Session and a Trace Recording Session according to the trace control and configuration parameters received in the IUR_INVOKE_TRACE message.

If there are insufficient resources in the DRNC, the DRNC may not start a Trace Recording Session.

The Trace Recording Session Reference sent by the SRNC to the DRNC shall be the same what SRNC has received in the CN_INVOKE_TRACE message from the CN.

Interaction with Relocation

If the tracing shall continue also after the relocation has been performed, the CN Invoke Trace procedure shall be re-initiated from the CN towards the future SRNC after the Relocation Resource Allocation procedure has been executed successfully.

4.2.3.2 PS Domain starting mechanisms

In SGSN/GGSN/BM-SC a Trace Recording Session should start after the reception of a Trace Session Activation message (in SGSN it is the MAP-ACTIVATE_TRACE_MODE, in GGSN it is the GTP-Create PDP Context request or Update PDP Context request, in BM-SC it is the Diameter Gmb AAR message) and if any of the defined start triggering events occur. During the Trace Recording Session, the SGSN/GGSN/BM-SC shall record the signalling messages in the interfaces that are defined in the list of interfaces parameter. The Trace Depth parameter defines whether entire signalling messages or just some IEs need to be recorded.

The SGSN/GGSN/BM-SC may not start a Trace Recording Session if there are insufficient resources available for the recording.

In case of an established session, the SGSN may start the Trace Recording Session immediately after the reception of the Trace Session Activation message. However, if any of the start triggering events occurs in SGSN after receiving the Trace Session activation message, it shall start the Trace Recording.

When a Trace Recording Session is started in SGSN, it shall assign a Trace Recording Session Reference for the Trace Recording Session. When the SGSN propagates the Trace control and configuration parameters to GGSN or to UTRAN (I.e. activates a Trace Session in GGSN/UTRAN), it shall include the assigned Trace Recording Session Reference in the Trace Session Activation message. When an SGSN starts a Trace Recording Session and the list of NE types parameter requires GGSN tracing, it shall send the GTP- Update PDP Context Request or Create PDP Context Request message for activating the Trace Session to GGSN. When a GGSN starts a Trace Recording Session and the list of NE types parameter requires BM-SC tracing, it shall send a Diameter Gmb AAR message to the BM-SC in order to activate a Trace Session in the BM-SC. Also, when an SGSN starts a Trace Recording Session and the list of NE types parameter requires RNC tracing, it shall send the CN_INVOKE_TRACE message to the RNC in order to activate a Trace Session in RNC. In both cases the Trace Session and the Trace Recording Session in the receiving NE should start at the same time.

In case of SRNS relocation the SGSN shall send the CN_INVOKE_TRACE message to the new SRNC after the successful Relocation Resource Allocation procedure.

SGSN has to find the identity of the mobile before it activates a Trace Session towards other NE. The IMEI(SV) can be got from the Mobile by using the Identification procedure on the Iu interface.

When the SGSN sends the Trace Session activation (CN_INVOKE_TRACE) message to RNC it shall include the following parameters to the message:

– Trace Target: IMSI or IMEI (SV) (M).

– Trace Reference (M).

– Trace Recording Session Reference (M).

– Trace Depth (M).

– List of Interfaces (O).

– Trace Collection Entity IP Address (O).

4.2.3.3 CS Domain starting mechanisms

In MSC Server/MGW a Trace Recording Session should start after the reception of a Trace Session Activation message (MAP-ACTIVATE TRACE MODE in MSC Server and ADD/MOD command with Trace package in MGW) and if any of the defined start triggering events occur. During the Trace Recording Session the MSC Server/MGW shall record the signalling messages in the interfaces that are defined in the list of interfaces parameter. The Trace Depth parameter defines whether entire signalling messages or just some IEs need to be recorded.

The MSC Server may not start a Trace Recording Session if there are insufficient resources available for the recording.

In case of an established call, the MSC Server may start the Trace Recording Session immediately after the reception of the MAP-ACTIVATE_TRACE_MODE message. However, if any of the start triggering events occur in the MSC Server after receiving the Trace Session activation message, it shall start the Trace Recording Session.

When a Trace Recording Session is started in MSC Server, it shall assign a Trace Recording Session Reference for the Trace Recording Session. When the MSC Server propagates the Trace control and configuration parameters to MGW or to UTRAN (I.e. activates a Trace Session in MGW/UTRAN) it shall include the assigned Trace Recording Session Reference in the Trace Session Activation message.

When an MSC Server starts a Trace Recording Session and the list of NE types parameter requires MGW tracing, it shall send the ADD/MOD command with trace package to MGW in order to activate the trace in MGW. Also, when an MSC Server starts a Trace Recording Session and the list of NE types parameter requires RNC tracing, it shall send the CN_INVOKE_TRACE message to the RNC. In both cases the Trace Session and the Trace Recording Session in the receiving NE should start at the same time.

MSC Server has to find the identity of the mobile before it activates a Trace Session towards other NE. The IMEI(SV) can be got from the Mobile by using the Identification procedure on the Iu interface.

In case of SRNS relocation the MSC Server shall send the CN_INVOKE_TRACE message to the new SRNC after the successful Relocation Resource Allocation procedure. The following figure shows an example how the Trace Session is activated with CN_INVOKE_TRACE message in case of relocation.

Figure 4.2.3.3.1: Starting a Trace Recording Session (Signalling) in CS Domain

When the new SRNC receives the CN_INVOKE_TRACE message it should start immediately a Trace Session and a Trace Recording session according to the trace control and configuration parameters received in the CN_INVOKE_TRACE message. The Trace Session shall automatically be deactivated in the old RNC when the Iu connection is released.

When the MSC Server sends the Trace Session activation (CN_INVOKE_TRACE) message to RNC it shall include the following parameters to the message:

– Trace Target: IMSI or IMEI (SV) (M).

– Trace Reference (M).

– Trace Recording Session Reference (M).

– Trace Depth (M).

– List of Interfaces to trace (O).

– Trace Collection Entity IP Address (O).

4.2.3.4 Void

4.2.3.5 Service level tracing for IMS starting mechanism

4.2.3.5.1 General

A trace recording session should start when there is an active trace session and when an appropriate start triggering event occurs. Figure 4.2.3.5.1.1 illustrates the initiation of a trace recording session at the UE, P-CSCF and S-CSCF within an originating network when any of the defined triggering events as defined in Trace Session Activation occur. When a triggering event occurs in the UE (Step 1) it includes in the outgoing SIP (service) signalling message a service level tracing Start Triggering Event (Step 2) and starts a trace recording session (Step 3). When the P-CSCF receives the SIP (service) signalling message containing the Start Triggering Event it authenticates the received Start Triggering Event (step 4) and starts its trace recording session (Step 5). The P-CSCF forwards the SIP (service) signalling message containing the Start Triggering Event to the S-CSCF (Step 7) and starts its trace recording session (Step 8).

Figure 4.2.3.5.1.1a: Starting a Trace Recording Session within originating network

Figure 4.2.3.5.1.2 illustrates the initiation of a trace recording session at the AS, I-CSCF, HSS, S-CSCF, P-CSCF and UE within the terminating network when any of the defined triggering events as defined in Trace Session Activation occur.

NOTE: All origination, termination and S-CSCF to CSCF procedures as described in 3GPP TS 23.228 [15] apply.

Figure 4.2.3.5.1.1b: Starting a Trace Recording Session within terminating network

When S-CSCF#1 receives the SIP (service) signalling message containing the Service Level Tracing Start Trigger Event (step 7) it shall start a trace recording session (Step 8). Based on the service control information (step 10) S-CSCF#1 forwards the SIP (service) signalling message containing the Start Trigger Event to the Application Server (Step 11).

On reception of the SIP INVITE the Application Server adds, removes, or modifies the header contents contained in the SIP INVITE (see 3GPP TS 23.218) and proxies the SIP INVITE together with the Start Trigger Event back to S-CSCF#1 (Step 14). The Application Server also starts a trace recording session (Step 12).

S-CSCF#1 forwards the SIP INVITE containing the service level tracing Start Trigger Event request to I-CSCF#2 (Step 15). At this point the I-CSCF starts a trace recording session Step 16).

I-CSCF#2 initiates a query to the HSS for the current location of the terminating user (UE#2) and includes in the Cx-User Location procedure the service level tracing Start Trigger Event (Step 20).

When the HSS receives the query for the current location and an associated Start Trigger Event it shall start a trace recording session (Step 19) and returns to the I-CSCF#2 the address of the current S-CSCF (S-CSCF#2) for the terminating user and the service level tracing Start Trigger Event (20).

I-CSCF#2 forwards the SIP INVITE containing the Start Trigger Event to the S-CSCF (S-CSCF#2) that will handle the session termination. When the S-CSCF receives the SIP INVITE containing the Start Trigger Event it starts a trace recording session (Step 21).

The S-CSCF forwards the SIP INVITE containing the Start Trigger Event to the P-CSCF (P-CSCF#2). When the P-CSCF receives the SIP INVITE containing the Start Trigger Event it starts a trace recording session (Steps 24 and 25).

The P-CSCF forwards the SIP INVITE containing the Start Trigger Event to the terminating UE (UE#2). When the terminating UE receives the SIP INVITE containing the Start Trigger Event it starts a trace recording session (Step 28).

The continuation of the termination procedures is as defined in 3GPP TS 23.228 [15].

4.2.3.5.2 Starting mechanism at the UE

For a UE that has an active trace session (see subclause 4.1.2.9.4) one or more trace recording session(s) (e.g. to allow the tracing for several different simultaneous services) shall be started when any of the defined triggering events occur at the UE, and when the condition(s) as defined by the trace control and configuration parameters within the received management operation occur.

A Trace recording session(s) may be initiated at an originating UE when:

1) The UE detects the initiation of the specified service to be traced. The service may be initiated either by the end user or by an application.

The triggering events at a terminating UE include:

1) The UE detects the initiation of the specified service to be traced. The service may be initiated either by the end user or by an application.

2) The UE detects the reception of an incoming SIP message containg the service level tracing Start Triggering Event.

A Trace recording session(s) may be initiated at a UE (both originating and terminating) when it detects a start trigger event initiated directly by the Device Management server for the purpose of allowing not only SIP information related to the service to be traced, but also information relating to the processes performed by the UE to support the initialization of the service.

Upon the detection of a triggering event the UE shall include in the appropriate outgoing SIP (service) signalling message (i.e. the outgoing signalling messages associated with the service to be traced) a service level tracing Start Triggering Event.

4.2.3.5.3 Starting mechanism at the IMS NE

For an IMS NE (i.e. S/I/P-CSCF, AS, HSS, MRF, MGCF, BGCF) that has an active trace session (see subclause 4.1.2.9) a trace recording session should be started when it receives in an incoming SIP (service) signalling message or DIAMETER signalling message containing a service level tracing Start Triggering Event and when the information contained within the service level tracing Start Triggering Event matches the information received by the IMS NE during trace session activation. The IMS NE shall also start the recording of signalling messages in the interfaces that are defined in the list of received interfaces parameter.

An IMS NE (i.e. S/I/P-CSCF, AS, HSS, MRF, MGCF, BGCF) that receives an incoming SIP (service) signalling message containing a service level tracing Start Triggering Event should forward in an appropriate outgoing SIP (service) signalling message (i.e. outgoing signalling messages associated with the service being traced) the same service level tracing Start Triggering Event (i.e. service level tracing Start Triggering Event with the same Trace Reference).

When an IMS NE has an active trace session and trace recording session, and when an incoming SIP (service) signalling message is part of an existing dialog or standalone transaction and contains a service level tracing Start Trigger Event the IMS NE shall determine that an active Trace Recording Session exists and shall not start a new Trace Recording Session.

Depending on operator policy, a HSS may forward the service level tracing Start Triggering Event to an external AS (see 3GPP TS 23.218 [14]). In the case of a terminating session a S-CSCF or I-CSCF may forward the service level tracing Start Triggering Event to a P-CSCF in a visited IM CN SS. A P-CSCF shall send a service level tracing Start Triggering Event to a terminating UE.

When a P-CSCF receives a SIP (service) signalling message containing a service level tracing Start Triggering Event from a UE it shall authenticate the Start Triggering Event by comparing the information contained within the received service level tracing Start Triggering Event (see subclause 5.2) either against the information it received within the Start Trace activation message or by requesting information from the I-CSCF or S-CSCF. If the received service level tracing Start Triggering Event is authenticated by the P-CSCF it should start a trace recording session and shall forward the service level tracing Start Triggering Event in the appropriate outgoing SIP (service) signalling message.

If the authentication of the incoming service level tracing Start Triggering Event fails the P-CSCF shall not start a trace recording session and shall not forward the service level tracing Start Triggering Event in any outgoing SIP (service) signalling message. The P-CSCF should provide an indication to the Management System following the unsuccessful authentication of the service level tracing Start Triggering Event.

When an IMS NE does not have an active trace session when it receives an incoming SIP (service) signalling message that contains a service level tracing Start Trigger Event, the IMS NE shall not initiate a Trace Recording Session and should forward in an appropriate outgoing SIP (service) signalling message the same service level tracing Start Trigger Event.

4.2.3.5.4 Charging concepts for Service Level Tracing for IMS

Charging for Service Level Tracing for IMS shall be fulfilled using IMS charging mechanism as specified in TS 32.240 [19] and TS 32.260 [20].

It shall be possible to apply specific tariffs (e.g. zero rating) to the bearer and/or signalling traffic associated with services subjected to Service Level Tracing for IMS.

As described in subclause 4.2.3.5 an IMS NE that has an active trace session should start a trace recording session when it detects a service level tracing Start Triggering Event. An IMS NE shall also provide an indication in the generated charging information that service level tracing has been applied.

4.2.3.6 E-UTRAN starting mechanism

In an eNB the Trace Recording Session will always be the same as the Trace Session as no triggering events are defined in eNB.

Tracing starts immediately at eNodeB upon reception of the trace control and configuration parameters. The eNodeB may not start a Trace Recording Session if there are insufficient resources available for the recording, however, the eNodeB shall store the trace control and configuration parameters, and forward these parameters when the UE handovers to other eNBs over X2.

The Trace Recording Session shall be started at the eNB when it receives trace control and configuration parameters via one of the following messages:

1) via an S1-Initial Context Setup Request message from the MME in response to an S1-Initial UE Message

2) via an S1-Trace Start message from the MME in response to an S1-Initial UE Message or when an established S1AP connection exists

3) via an S1-Handover Request message from the target MME as part of intra/inter-MME handover procedures via S1

4) via an X2-Handover Request message from a source eNodeB as part of inter-eNodeB handover procedures via X2

There can only be one Trace Recording Session Reference per Trace Reference at one given time for a UE trace session. So there shall be only one TR/TRSR to be propagated during S1 and X2 handover.

If the Trace Reference is the same as an existing Trace Session for the same subscriber or equipment, and the Trace Recording Session Reference is the same as the existing Trace Recording Session in the existing Trace Session having the same Trace Reference, the eNB shall not start a new Trace Recording Session and shall continue with the existing trace session and ignore the second request.

If the Trace Reference is the same as an existing Trace Session for the same subscriber or equipment, and the Trace Recording Session Reference is not the same as the existing Trace Recording Session in the existing Trace Session having the same Trace Reference, the eNB shall continue with the existing trace session and ignore the second request.

4.2.3.7 EPC starting mechanisms

In MME/SGW/PGW a Trace Recording Session should start after the reception of a Trace Session Activation message and if any of the defined start triggering events occur. During the Trace Recording Session, the MME/SGW/PGW shall record the signalling messages in the interfaces that are defined in the list of interfaces parameter. The Trace Depth parameter defines whether entire signalling messages or just some IEs need to be recorded.

The MME/SGW/PGW may not start a Trace Recording Session if there are insufficient resources available for the recording.

In case of an established session, the MME/SGW/PGW may start the Trace Recording Session immediately after the reception of the trace control and configuration parameters. However, if any of the start triggering events occurs in MME/SGW/PGW after receiving the trace control and configuration parameters, it shall start the Trace Recording Session.

In the case of the triggering events come into collision on the same traced UE as defined in 3GPP TS 24.301[33], the MME shall not start a new Trace Recording Session for the later event(s), and shall use the existing Trace Recording Session and Trace Recording Session Reference to continuing the trace recording for these events until one stop triggerring event occurs.

MME shall start a Trace Recording Session for a certain Trace Session only if there is no ongoing Trace Recording Session for this Trace Session. i.e. at any given time, there can be a maximum of one Trace Recording Session for a certain Trace Session.

When a Trace Recording Session is started in MME, it shall assign a Trace Recording Session Reference for the Trace Recording Session. When the MME propagates the Trace control and configuration parameters to E-UTRAN (i.e. activates a Trace Session in eNB), it shall include the assigned Trace Recording Session Reference in the Trace Session Activation message.

Also, when an MME starts a Trace Recording Session and the list of NE types parameter requires eNB tracing, it shall propagate the trace control and configuration parameters including the Trace Recording Session Reference via the S1 interface to the eNodeB per one of the following messages:

1) if an S1 connection exists, via the S1-Trace Start message

2) if the S1 connection does not exist, via the S1-Trace Start message prior to S1 connection setup, or via the S1-Initial Context Setup Request message during S1 connection setup

3) during intra/inter-MME handover over S1, via the S1-Handover Request message

In above cases the Trace Session and the Trace Recording Session in the receiving NE should start at the same time

If all events are set in the triggering event parameter at the MME, MME shall send Trace Session Activation message to eNB not only when the MME starts the Trace Recording Session, but also when an Intra-MME handover happens. In this case the MME shall send Trace Session activation to the target eNB via the S1-Handover Request message..

NOTE: In case of "UE-Initiated Detach Procedure with UE camping on GERAN/UTRAN and ISR activated / SGSN-Initiated Detach Procedure with ISR activated",
Trace is not activated in eNB.

4.2.3.8 EPC starting mechanisms for MDT

In the MME, no trace recording sessions are started for MDT trace sessions. The MME sends the trace session activation to the eNodeB with parameters as specified in 4.1.2.12.

4.2.3.9 E-UTRAN starting mechanisms for MDT

A trace recording session of either immediate or logged MDT or Logged MBSFN MDT shall be started in the eNodeB for a given UE when a trace session activation request is received from the MME for the UE and the MDT UE selection conditions are satisfied for the UE. The eNodeB shall configure the corresponding MDT RRC measurements at the UE. If selection conditions are not satisfied, the eNodeB shall store the trace control and configuration parameters, and forward these parameters when the UE handovers to other eNBs over X2 or S1.

If the eNodeB receives a Signalling Based MDT activation request when the UE is served by a cell that is in the eNodeB but not in the MDT area scope then the eNodeB shall store the MDT configuration and configure the UE when the UE moves to a cell in the eNodeB (intra eNodeB handover) that satisfies the area scope in the request.

4.2.3.10 Starting mechanisms at UE for MDT

There is no starting mechanism at the UE for MDT trace recording sessions. The UE shall execute the received MDT measurement configuration. In case of logged MDT or Logged MBSFN MDT,, the UE shall store the trace recording session parameters as received from the eNodeB.

4.2.3.11 5GC starting mechanisms

In AMF, PCF, SMF or UPF a Trace Recording Session should start after the reception of a Trace Session Activation message and if any of the defined start triggering events occur. During the Trace Recording Session, the AMF, PCF, SMF or UPF shall record the signalling messages in the interfaces that are defined in the list of interfaces parameter. The Trace Depth parameter defines whether entire signalling messages or just some IEs need to be recorded.

The AMF, PCF, SMF or UPF may not start a Trace Recording Session if there are insufficient resources available for the recording.

In case of an established session, the AMF, PCF, SMF or UPF may start the Trace Recording Session immediately after the reception of the trace control and configuration parameters. However, if any of the start triggering events occurs in AMF, PCF, SMF or UPF after receiving the trace control and configuration parameters, it shall start the Trace Recording Session.

In the case of the triggering events come into collision on the same traced UE as defined in 3GPP TS 24.501[45], the AMF shall not start a new Trace Recording Session for the later event(s), and shall use the existing Trace Recording Session and Trace Recording Session Reference to continuing the trace recording for these events until one stop triggerring event occurs.

AMF shall start a Trace Recording Session for a certain Trace Session only if there is no ongoing Trace Recording Session for this Trace Session. i.e. at any given time, there can be a maximum of one Trace Recording Session for a certain Trace Session.

When a Trace Recording Session is started in AMF, it shall assign a Trace Recording Session Reference for the Trace Recording Session. When the AMF propagates the Trace control and configuration parameters to NG-RAN (i.e. activates a Trace Session in NG-RAN node), it shall include the assigned Trace Recording Session Reference in the Trace Session Activation message.

Also, when an AMF starts a Trace Recording Session and the list of NE types parameter requires NF-RAN node tracing, it shall propagate the trace control and configuration parameters including the Trace Recording Session Reference via the NG interface to the NG-RAN node per one of the following messages:

1) if an NG connection exists, via the NG-Trace Start message

2) if the NG connection does not exist, via the NG-Trace Start message prior to NG connection setup, or via the NG-Initial Context Setup Request message during NG connection setup

3) during intra/inter-AMF handover over NG, via the NG-Handover Request message

In above cases the Trace Session and the Trace Recording Session in the receiving NE should start at the same time

If all events are set in the triggering event parameter at the AMF, AMF shall send Trace Session Activation message to NG-RAN node not only when the AMF starts the Trace Recording Session, but also when an Intra-AMF handover happens. In this case the AMF shall send Trace Session activation to the target NG-RAN node via the NG-Handover Request message.

4.2.3.12 NG-RAN starting mechanism

In a NG-RAN node the Trace Recording Session will always be the same as the Trace Session as no triggering events are defined in NG-RAN node.

Tracing starts immediately at NG-RAN node upon reception of the trace control and configuration parameters Trace Activation IE). The NG-RAN node may not start a Trace Recording Session if there are insufficient resources available for the recording, however, the NG-RAN node shall store the trace control and configuration parameters, and forward these parameters when the UE handovers to other NG-RAN nodes over Xn or when other NG-RAN node retrieves the UE Context over Xn.

The Trace Recording Session shall be started at the NG-RAN node when it receives trace control and configuration parameters via one of the following messages:

1) via an INITIAL CONTEXT SETUP REQUEST message over NG from the AMF

2) via a TRACE START message over NG from the AMF. If no UE-associated logical NG-connection exists, the UE-associated logical NG-connection will be established as part of the Trace Start procedure.

3) via an NG HANDOVER REQUEST message from the target AMF as part of intra/inter-AMF handover procedures via NG.

4) via an Xn HANDOVER REQUEST message from a source NG-RAN node as part of inter-NG-RAN node handover procedures via Xn

5) via an Xn RETRIEVE UE CONTEXT RESPONSE message from old NG-RAN node after been sent to RRC_INCTIVE and then establishing RRC Connection to a new NG-RAN node

There can only be one Trace Recording Session Reference per Trace Reference at one given time for a UE trace session. So there shall be only one TR/TRSR to be propagated during NG and Xn handover.

If the Trace Reference is the same as an existing Trace Session for the same subscriber or equipment, and the Trace Recording Session Reference is the same as the existing Trace Recording Session in the existing Trace Session having the same Trace Reference, the NG-RAN node shall not start a new Trace Recording Session and shall continue with the existing trace session and ignore the second request.

If the Trace Reference is the same as an existing Trace Session for the same subscriber or equipment, and the Trace Recording Session Reference is not the same as the existing Trace Recording Session in the existing Trace Session having the same Trace Reference, the NG-RAN node shall continue with the existing trace session and ignore the second request.

4.2.3.13 5GC starting mechanisms for signalling based MDT

There is no starting mechanism in 5GC for signalling based MDT as no trace recording sessions are started for MDT trace sessions in AMF.

4.2.3.14 NG-RAN starting mechanisms for signalling based MDT

A trace recording session of either immediate or logged MDT shall be started in the gNB for a given UE when a trace session activation request is received from the AMF for the UE and the MDT UE selection conditions are satisfied for the UE. The gNB shall configure the corresponding MDT RRC measurements at the UE. If selection conditions are not satisfied, the gNB shall store the trace control and configuration parameters and forward these parameters when the UE handovers to other gNBs over Xn or N2.

If the gNB receives a signalling based MDT activation request when the UE is served by a cell that is in the gNB but not in the MDT area scope then the gNB shall store the MDT configuration and configure the UE when the UE moves to a cell in the gNB (intra gNB handover) that satisfies the area scope in the request.