hash
stringlengths 32
32
| doc_id
stringlengths 5
12
| section
stringlengths 4
595
| content
stringlengths 0
6.67M
|
|---|---|---|---|
556719e8696c25b4ac3572453768229c | 22.843 | 5.7.5 Existing features partly or fully covering the use case functionality | Identification of UAVs which don’t have subscription with the 5G operator for DAA has been specified in Rel-18. Tracking for such UAVs is currently not covered in Rel-18 because release 18 only supports the capability to allow the 5G operator tracking of those UAVs which have the subscription with the 5G operator. |
556719e8696c25b4ac3572453768229c | 22.843 | 5.7.6 Potential New Requirements needed to support the use case | [P.R 5.7.6-001] The 5G system shall be able to support a mechanism to enable a network operator to track a UAV which doesn’t have subscription with this network operator. |
556719e8696c25b4ac3572453768229c | 22.843 | 5.8 Use case on UAV simultaneous traffic over two networks | |
556719e8696c25b4ac3572453768229c | 22.843 | 5.8.1 Description | Reliability of UAV traffic is a very important challenge and target for the aviation industry, e.g., for command and control (C2) or other critical data. One scenario of interest includes UAVs, provided with dual SIM/subscription, able to connect simultaneously via two networks for transferring different type of traffic, e.g., using the most reliable network for C2 or other critical data, while other type of traffic is transmitted over the second network.
As an example, let us assume the case of a UAV company (“SafeSky”) that offers monitoring and surveillance services for specific venues (e.g., stadiums) or public events. Their drones are 5G-capable and configurable with dual subscription, one for nationwide PLMN connectivity (with MNO-A), one for ad-hoc UAV coverage using networks deployed in specific public safety venues/locations (can be a PLMN or NPN, managed by MNO-A or other MNO).
In such scenario, when there is available coverage from both NWs, SafeSky can configure their UAVs to transfer C2 traffic with a local UAV-Controller (UAV-C) via the ad-hoc PLMN1/NPN (most suitable for low volume/data rate, but very stringent real-time latency and reliability), while less critical user data is exchanged with the UAV Supplier Service (USS) and UTM via the wide-area PLMN2 coverage. This is illustrated, at high level, in Figure 5.8.1-1, where PLMN1/NPN is the local NW and PLMN2 is the wide area network.
Figure 5.8.1-1: Example of separating UAV traffic over two PLMNs
In other examples, different traffic differentiation configurations are possible (in a similar venue scenario), e.g., UAVs may transfer real-time critical payload traffic (e.g., streaming a 4K video of a sport event) with a local endpoint via the ad-hoc PLMN1/NPN, while C2 and UAV monitoring data via the wide-area PLMN2 network. |
556719e8696c25b4ac3572453768229c | 22.843 | 5.8.2 Pre-conditions | SafeSky UAV has dual subscription (dual SIM, or SIM and NPN credentials) to two different PLMNs (or a PLMN and an NPN). It is assumed that SafeSky has negotiated with both MNOs proper data policies for traffic generated by its UAVs over their respective networks (e.g., QoS, traffic profile/activity, charging, etc.).
UAV is capable of dual radio operation (e.g., NR+NR).
UAV’s application configuration policies include when/where to use connectivity via two PLMNs, which network to use for which traffic, specific C2 traffic/QoS parameters, etc. In this example, C2 traffic should go over PLMN1/NPN1 (offering ultra-reliable connectivity), other traffic over PLMN2.
It is assumed, in this use case, that UAV connects to the same USS/UTM and UAV-C during the whole flight operation.
This scenario assumes mostly UAV-originated traffic and/or that UTM is aware of which data goes over each PLMN connection, so that corresponding uplink and downlink traffic (e.g., for C2 communication) is transferred over the same PLMN connection (e.g., PLMN1). |
556719e8696c25b4ac3572453768229c | 22.843 | 5.8.3 Service Flows | Based on UAV configuration policy:
1. SafeSky UAV registers to both PLMN1/NPN1 and PLMN2, using their respective home subscription.
2. When monitoring surveillance service is started, UAV connects with both networks.
3. UAV traffic is transferred simultaneously on both networks, where C2 communication with the UAV-C is routed over PLMN1 and other traffic monitoring data with USS/UTM is routed over PLMN2. |
556719e8696c25b4ac3572453768229c | 22.843 | 5.8.4 Post-conditions | The surveillance service runs smoothly, with no glitch in UAV operation and manoeuvring. |
556719e8696c25b4ac3572453768229c | 22.843 | 5.8.5 Existing features partly or fully covering the use case functionality | There are no service requirements in TS 22.125 [3] about UAS multi-NW/PLMN support. Other requirements (e.g., from TS 22.261 [4] or TS 22.101 [5]) do not cover the specific target UAV scenarios and functionalities. Some examples of existing requirements are captured below:
[from TS 22.261 [4] sec. 6.18: Multi-network connectivity and service delivery across operators]
The 5G system shall enable users to obtain services from more than one network simultaneously on an on-demand basis.
For a user with a single operator subscription, the use of multiple serving networks operated by different operators shall be under the control of the home operator.
When a service is offered by multiple operators, the 5G system shall be able to maintain service continuity with minimum service interruption when the serving network is changed to a different serving network operated by a different operator.
NOTE 1: A business agreement is required between the network operators.
In the event of the same service being offered by multiple operators, unless directed by the home operator's network, the UE shall be prioritized to receive subscribed services from the home operator's network.
NOTE 2: If the service is unavailable (e.g., due to lack of network coverage) from the home operator's network, the UE may be able to receive the service from another operator's network.
NOTE 3: QoS provided by the partner operator's network for the same service will be based on the agreement between the two operators and could be different than that provided by the home operator's network.
[from TS 22.101 [5] sec. 13.4]
The 3GPP system shall support ME with multiple USIMs (on the same UICC or on different UICCs) that are registered at the same time.
The 3GPP system shall treat each registration from the USIMs of a MUSIM UE independently. Each registered USIM in a MUSIM UE shall be associated with a dedicated IMEI/PEI.
[from others]
In addition, there are existing UAV requirements (from TS 22.125 [3]) on 3GPP system to provide monitoring and notification regarding C2 communication changes, which are not specific to inter-PLMN connectivity, for example:
The 3GPP system shall support C2 communication with required QoS for pre-defined C2 communication models (e.g. using direct ProSe Communication between UAV and the UAV controller, UTM-navigated C2 communication based on flight plan between UTM and the UAV).
The 3GPP system shall support C2 communication with required QoS when switching between the C2 communication models.
The 3GPP system shall support a mechanism for the UTM to request monitoring of the C2 communication with required QoS for pre-defined C2 communication models (e.g. using direct ProSe Communication between UAV and the UAV controller, UTM-navigated C2 communication between UTM and the UAV). |
556719e8696c25b4ac3572453768229c | 22.843 | 5.8.6 Potential New Requirements needed to support the use case | [PR.5.8-001] The 5G system shall be able to support service enablement layer exposure mechanisms for the UTM or other authorized 3rd party to provide the UAV application with configuration information to route different traffic across different PLMN connections simultaneously, e.g. C2 traffic via one PLMN and other data via the second PLMN.
NOTE 1: The above requirements can be extended to scenarios where one network is a PLMN and one is an NPN.
NOTE 2: There is no impact on legacy network selection.
NOTE 3: It is assumed that UAV traffic handling, over each PLMN, is subject to network control mechanisms (e.g. in accordance with MNO traffic routing priorities, available QoS/resources, etc.). |
556719e8696c25b4ac3572453768229c | 22.843 | 5.9 Use case on UAV traffic over alternative networks | |
556719e8696c25b4ac3572453768229c | 22.843 | 5.9.1 Description | One of the Aviation industry recommendations for UAV operation (and AAM - Advanced Air Mobility) is the ability to support robust reliability and flexible redundancy of critical communication links, including cellular, e.g., for Command and Control (C2) and other flight-critical communications.
Within the context of improving 5G reliability and redundancy for UAVs, scenarios with dual subscription and multi-NW connectivity become of interest. For example, drones can be equipped with dual SIM (or dual credentials) and use one subscription for active communication with one network while the second subscription for setting up a “stand-by” connection path, to be used in case the first network becomes unsuitable or unavailable.
Figure 5.9.1-1 shows, at high level, the two connectivity options, where UAV traffic is first active on the preferred network (PLMN-1), then switched to the second network (PLMN-2). Other options are not precluded, e.g., one of the two networks could be an NPN.
Figure 5.9.1-1: UAV traffic over two alternative PLMNs (one active at a time)
Which PLMN to use, and when to switch between them, would be based on specific conditions and policies configurable by the UAV owner/operator.
One example is about a UAV operator (“SkyTV”), offering professional aerial video/TV services, in an environment where there is 5G coverage provided by PLMN1 (possibly shared with other PLMNs) overlapping with 5G coverage provided by PLMN2. It can be that both PLMNs are managed by the same MNO (MNO-A), or different MNOs, using different SIM/subscription credentials to access each PLMN. In order to support seamless aerial service and UAV traffic management, the whole flight path can require consistent network connection and complete network coverage. In certain areas, though, the coverage of the first PLMN cannot be guaranteed, or is not preferable (e.g. due to network congestion), while the coverage of PLMN2 is better, or preferred.
In such environment, SkyTV is contracted by a national TV broadcaster for video streaming of e.g., outdoor events, races, marathons etc. To improve the reliability and QoS experience, SkyTV can equip its UAVs with dual SIM subscription (to PLMN1 and PLMN2) and configure them such that PLMN-1 is the preferred network, while PLMN-2 should be the stand-by network, together with policies and conditions for switching between them. The example is further elaborated below. |
556719e8696c25b4ac3572453768229c | 22.843 | 5.9.2 Pre-conditions | SkyTV UAV has dual SIM subscription to PLMN1 and PLMN2. It is assumed that SkyTV has negotiated with both MNOs proper data subscriptions for traffic generated by its UAVs (e.g., QoS, traffic profile/activity, charging, etc.).
UAV’s application is configured to use both PLMNs via both subscriptions (USIM1/USIM2), transfer user data via PLMN-1 if/when available or suitable, otherwise switch to PLMN-2. Network switch is decided and controlled by UAV or UTM, based on configured application policies in the UAV or UTM.
UAV’s application policy can include e.g., when/where to use connectivity via two PLMNs, which PLMN should be active/stand-by, specific C2 traffic/QoS policies, switch conditions, etc.
It is assumed, in this use case, that UAV’s application connects to the same USS/UTM or UAV-C during the whole flight operation (regardless of which PLMN connection is used).
This scenario assumes mostly UAV-originated traffic and/or that UTM is aware of which PLMN connection is used by the UAV, so that both uplink and downlink traffic are transferred over the same active PLMN connection, and USS/UTM interacts with the active PLMN e.g., for UAS identification and traffic management. For example, when UAV application sends the data packets over one PLMN connection, the UTM can reply over the same PLMN connection. |
556719e8696c25b4ac3572453768229c | 22.843 | 5.9.3 Service Flows | 1. An outdoor marathon event is starting, and SkyTV UAV is set to follow the peloton for all the itinerary (20 miles). SkyTV service provider has configured the UAV application with specific policy on how to route traffic over the two subscribed PLMN networks.
2. Before the kick-off, UAV’s communication module using USIM1 and USIM2 is registered to both PLMNs, has established C2 connection with the UAV-C or USS/UTM, using respective subscriptions/USIM credentials, and is idle (no traffic). Based on UAV operator (SkyTV) policy, PLMN-1 is configured initially as active network (to transfer video data and C2 data) while PLMN-2 is configured as standby network.
2. At the start, UAV’s application transfers real-time video data and exchange control and command data with the USS/UTM or UAV-C via PLMN1, while PLMN-2 connection remains inactive (standby connection).
3. Based on configured UAV application policies, along the itinerary, UAV video traffic and C2 data is switched over PLMN-2: UAV switches its active C2 connection with UAV-C or USS/UTM to PLMN2 while PLMN-1 becomes the standby network.
4. Later, video and C2 traffic is switched back by UAV to PLMN-1 (PLMN2 is inactive).
Steps 3&4 can be repeated along the route. |
556719e8696c25b4ac3572453768229c | 22.843 | 5.9.4 Post-conditions | SkyTV video service runs smoothly for all 20 miles, with no service degradation when transitioning in and out of PLMN-1 spotty coverage areas and the UAV is under control and monitored by USS/UTM and UAV-C all the time. |
556719e8696c25b4ac3572453768229c | 22.843 | 5.9.5 Existing features partly or fully covering the use case functionality | There are no service requirements in TS 22.125 [3] about UAS multi-NW/PLMN support. Other requirements (e.g., from TS 22.261 [4] or TS 22.101 [5]) do not cover the specific target UAV scenarios and functionalities. Some examples of existing requirements are captured below:
[from TS 22.261 [4] sec. 6.18: Multi-network connectivity and service delivery across operators]
The 5G system shall enable users to obtain services from more than one network simultaneously on an on-demand basis.
For a user with a single operator subscription, the use of multiple serving networks operated by different operators shall be under the control of the home operator.
When a service is offered by multiple operators, the 5G system shall be able to maintain service continuity with minimum service interruption when the serving network is changed to a different serving network operated by a different operator.
NOTE 1: A business agreement is required between the network operators.
In the event of the same service being offered by multiple operators, unless directed by the home operator's network, the UE shall be prioritized to receive subscribed services from the home operator's network.
NOTE 2: If the service is unavailable (e.g., due to lack of network coverage) from the home operator's network, the UE may be able to receive the service from another operator's network.
NOTE 3: QoS provided by the partner operator's network for the same service will be based on the agreement between the two operators and could be different than that provided by the home operator's network.
[from TS 22.101 [5] sec. 13.4]
The 3GPP system shall support ME with multiple USIMs (on the same UICC or on different UICCs) that are registered at the same time.
The 3GPP system shall treat each registration from the USIMs of a MUSIM UE independently. Each registered USIM in a MUSIM UE shall be associated with a dedicated IMEI/PEI.
[from others]
In addition, there are existing UAV requirements (from TS 22.125 [3]) on 3GPP system to provide monitoring and notification regarding C2 communication changes, which are not specific to inter-PLMN connectivity, for example:
The 3GPP system shall support C2 communication with required QoS for pre-defined C2 communication models (e.g., using direct ProSe Communication between UAV and the UAV controller, UTM-navigated C2 communication based on flight plan between UTM and the UAV).
The 3GPP system shall support C2 communication with required QoS when switching between the C2 communication models.
The 3GPP system shall support a mechanism for the UTM to request monitoring of the C2 communication with required QoS for pre-defined C2 communication models (e.g., using direct ProSe Communication between UAV and the UAV controller, UTM-navigated C2 communication between UTM and the UAV). |
556719e8696c25b4ac3572453768229c | 22.843 | 5.9.6 Potential New Requirements needed to support the use case | [PR 5.9.6-001] The 5G system shall be able to support service enablement layer exposure mechanisms for the UTM or other authorized 3rd party to provide the UAV application with configuration information to route and switch traffic between one active and one standby PLMN connection, e.g. for C2 communication reliability and redundancy purpose.
NOTE 1: The above requirement can be extended to scenarios where one network is a PLMN and one is an NPN.
NOTE 2: There is no impact on legacy network selection.
NOTE 3: It is assumed that UAV traffic handling, over each PLMN, is subject to NW control mechanisms (e.g. in accordance with MNO routing priorities, available QoS/NW resources, etc.). |
556719e8696c25b4ac3572453768229c | 22.843 | 5.10 Use case on UAV control in different aerial flight zones | |
556719e8696c25b4ac3572453768229c | 22.843 | 5.10.1 Description | 5G networks can provide connectivity to UAV for a variety of services, e.g., command and control, telemetry, remote ID, DAA, payload transfer (e.g., video), etc. In some scenarios, it would be beneficial to control and differentiate UAV operation and communication, e.g., regarding the use of specific network resources and/or specific application(s) settings, in different geographical areas. Some aspects of the 5G connectivity configuration, indeed, may need to be different depending on the area where the UAV is flying, e.g., in urban versus rural, or in areas close to high-risk obstacles or with stricter safety restrictions, like airports, stadiums, etc. This use case describes some means to efficiently support and control differentiation of UAV traffic across different zones and corresponding potential requirements.
For example, the UAV and UTM provider may identify different types of aerial flight zones in the territory, e.g., based on data traffic information, flight statistics, network topology. local conditions, events, etc. Based on business agreement with an MNO, the UTM and network can configure specific information on different flight zones, which can include e.g., the geographical location (footprint or 3D volume) of the flight zones and their mapping to different network areas, associated zone type or other means to identify a specific flight zone, and target UAV communication/QoS parameters (per flight zone).
UTM and/or the network can then convey to the UAV corresponding policy configuration (static or dynamic), controlling how the UAV should use connectivity services for various applications in different flight zones (e.g., per zone type). The network should also assist UAV to identify and differentiate specific flight zones, so that a UAV can apply the right configured policy based on the actual flight zone it is currently located in.
A general example of flight zones settings is illustrated in Fig.1, based on different geographical, logistical and safety factors (e.g., around city centre, stadiums, airports, harbour, rural areas, etc.). The figure may represent the flight zone settings for a specific day, or time of the day, and may vary based on local/temporary change of conditions.
Figure 5.10.1-1: Example of UAV flight zones configuration
The UAV policy could also contain a default configuration that can be used when no specific flight zone type is applicable or not indicated/available to the UAV.
The above would provide the ability for UTM to collaborate with the 5G network to provide flexible control of UAV communication over the identified flight zones, including scalable and dynamic adaptation based on temporary/local changes (e.g., data traffic, NW resources, environmental changes, local events or restrictions), by creating new flight zones that map on specific network areas and/or modifying existing zone types, etc.
One specific scenario and flow is further elaborated below. |
556719e8696c25b4ac3572453768229c | 22.843 | 5.10.2 Pre-conditions | A UAV service provider (U-Serv), offering monitoring & surveillance UAV services in multiple cities of the country, has an agreement with PLMN-A operator, including the configuration of two different flight zones (for all cities) mapped to different parts of the territory and network, with the following UAV application/traffic policies and conditions:
• - Zone A (no restriction/condition): for most areas/location of the cities’ territory
○ normal/default UAV application and communication settings
• - Zone B (with restriction/condition): for areas/locations e.g., with larger data traffic/congestion and/or higher drones’ density (e.g., where/when special events or emergency situations happen).
○ Increased frequency and QoS of C2 communication (for more reliable command and control operation)
○ Reduced data traffic activity/QoS for other communications (to reduce network congestion)
The Zone B policy configuration may also include other options, e.g., restrictions/conditions may vary based on UAV altitude (above / below a certain threshold, or within a certain range), time of the day, etc. |
556719e8696c25b4ac3572453768229c | 22.843 | 5.10.3 Service Flows | 1. 1. A U-Serv UAV, located in Zone A, is registered to PLMN-A and starts its daily surveillance service.
2. 2. The UAV detects it is in Zone A, e.g., if indicated by the network or based on UAV available information or pre-configuration, and connects to the network using the corresponding traffic policy for Zone A (normal/default operation).
3. 3. Along the UAV itinerary, UAV detects or is notified of being in a new zone type, Zone B, thus UAV application will apply corresponding traffic settings according to Zone B policy.
4. 4. If UAV detects or is notified that flight zone type changes to Zone A, UAV application will switch back to default traffic policy.
Note that change of flight zone type may happen due to the UAV moving across different geographical areas (associated to different zone types) or by change of zone type for the same UAV geographical location (e.g., based on temporary events). |
556719e8696c25b4ac3572453768229c | 22.843 | 5.10.4 Post-conditions | U-Serv UAV application can provide smooth service and is able to adapt its communication settings when flying across special locations/areas (zone B), based on UAV traffic policies and flight zone settings under control of UTM and network. |
556719e8696c25b4ac3572453768229c | 22.843 | 5.10.5 Existing features partly or fully covering the use case functionality | None identified. |
556719e8696c25b4ac3572453768229c | 22.843 | 5.10.6 Potential New Requirements needed to support the use case | [PR 5.10.6-001] The 5G system shall be able to support mechanisms for the UTM to configure different aerial flight zones where UAV application settings and communication QoS may be different, and provide network and UAV with means to identify those flight zones.
[PR 5.10.6-002] The 5G system shall be able to support mechanisms for the UTM to provide network and UAV with policy information including UAV application settings and communication QoS to be applied in specific aerial flight zones, e.g. based on flight zone type indicated or available to the UAV in a certain geographical location. |
556719e8696c25b4ac3572453768229c | 22.843 | 6 Consolidated potential requirements | |
556719e8696c25b4ac3572453768229c | 22.843 | 6.1 Network support and exposure for UAV usage | CPR #
Consolidated Potential Requirement
Original PR #
Comment
CPR 6.1-001
Based on operator’s policy, the 5G system shall be able to support a method to predict, monitor network conditions and QoS (e.g. bitrate, latency, reliability) and report to 3rd party along a continuous geographic planned flight path of a UAV at specific times of its expected flight duration.
P.R 5.2.6-001
P.R 5.6.6-001
CPR 6.1-002
Based on operator’s policy, the 5G system shall be able to provide UTM with the information about geographic areas where UAV service requirements could or could not be met based on predicted network conditions and QoS (e.g. bitrate, latency, reliability).
P.R 5.2.6-002
CPR 6.1-003
Based on a 3rd party request, the 5G system shall be able to assist the UTM with mechanism to provide to the 3rd party alternative UAV flight paths, e.g.,based on required waypoints, QoS, and exclusion zones.
P.R 5.2.6-003
CPR 6.1-004
Based on a 3rd party request and operator’s policy, the 5G system shall be able to reconfigure network resources to provide the required QoS along a UAV planned flight path, e.g. at particular geographical area(s) and time(s).
P.R 5.2.6-004
CPR 6.1-005
The 5G system shall be able to support a mechanism to enable a network operator to track an airborne connected UE which does not have an aerial subscription .
P.R 5.7.6-001
CPR 6.1-006
The 5G system shall be able to support service enablement layer exposure mechanisms for the UTM or other authorized 3rd party to provide the UAV application with configuration information to route and switch traffic between one active and one standby PLMN connection, e.g. for C2 communication reliability and redundancy purpose, or to route different traffic across different PLMN connections simultaneously, e.g. C2 traffic via one PLMN and other data via the second PLMN.
NOTE 1: The above requirement can be extended to scenarios where one network is a PLMN and one is an NPN.
NOTE 2: There is no impact on legacy network selection.
NOTE 3: It is assumed that UAV traffic handling, over each PLMN, is subject to NW control mechanisms (e.g. in accordance with MNO routing priorities, available QoS/NW resources, etc.).
P.R 5.8.6-001
P.R 5.9.6-001
CPR 6.1-007
The 5G system shall be able to support mechanisms for the UTM to configure different aerial flight zones where UAV application settings and communication QoS may be different, and provide network and UAV with means to identify those flight zones.
P.R 5.10.6-001
CPR 6.1-008
The 5G system shall be able to support mechanisms for the UTM to provide network and UAV with policy information including UAV application settings and communication QoS to be applied in specific aerial flight zones, e.g. based on flight zone type indicated or available to the UAV in a certain geographical location.
P.R 5.10.6-002
CPR 6.1-009
Based on operator and UTM policy, the 5G system shall be able to provide UTM with an airborne UE’s location and 3GPP identity to fulfil the UTM’s request.
NOTE 6: The 3GPP identity and UE location is expected to be used by the UTM to determine the position of airborne objects containing the UE.
P.R 5.5.6-001
CPR 6.1-010
Subject to user consent and national or regional regulation, based on operator and UTM policy, and the UTM’s request, the 5G system may be able to provide aerial object location(s) derived using 5G Wireless Sensing to the UTM to fulfil the UTM’s request.
NOTE 4: The airborne object location(s) are expected to be used by the UTM to determine the position of airborne objects.
P.R 5.5.6-002 |
556719e8696c25b4ac3572453768229c | 22.843 | 6.2 Safety and Security | CPR #
Consolidated Potential Requirement
Original PR #
Comment
CPR 6.2-001
The 5G system shall be able to detect that a connected UE is airborne, when the UE doesn’t have an aerial subscription..
PR 5.1.6-001
CPR 6.2-002
The 5G system shall support means to determine whether UTM supports regulatory requirements e.g., maximal VLOS distance between the UAV and the UAV controller.
NOTE 1: The requirement applies to the case of both UAV and UAV controller connected to the 3GPP network
PR 5.3.6-001
CPR 6.2-003
Based on operator’s policy, the 5G system shall be able to support a method to monitor and provide a 3rd party with information about deviations and violations along a UAV flight path and time.
NOTE 2: Deviations can be e.g. in location and/or time with respect to the original flight plan. Violations can be e.g. with respect to exclusion zones provided together with the flight plan or known via other means.
P.R 5.6.6-002
CPR 6.2-004
Based on operator’s policy, the 5G system shall be able to support a method to provide UTM and UAVs with the information collected or generated by the 5G system (e.g., based on sensing results), including e.g. the location or relative distance between3GPP UAVs and other flying objects (can be drones not using 3GPP connectivity).
P.R 5.4.6-001
CPR 6.2-005
The 5G system shall be able to track the UAV-Controller, regardless of the type of connection.
PR 5.3.6-002
CPR 6.2-006
Based on MNO policies and/or regulatory requirements, the 5G system shall be able to inform the UTM, when the 5G system detects a specific UAV condition or event, in order to enable UTM control of the UAV communication (e.g., when detecting violation of exclusion zones or maximal distance between UAVs).
PR 5.3.6-003 |
556719e8696c25b4ac3572453768229c | 22.843 | 7 Conclusion and recommendations | The present document contains several use cases and service requirements for enhanced UAV usage and safety. It is recommended to define normative service requirements based on the consolidated potential requirements identified in Section 6. Annex A (informative): Change history Change history Date Meeting TDoc CR Rev Cat Subject/Comment New version 2022-08 SA1#99-e S1-222402 TR skeleton 0.0.0 2022-09 SA1#99-e S1-222403, S1-222404, S1-222405, S1-222406 Scope Overview New use case - UAV detection New use case_Supporting UAV pre-flight preparation 0.1.0 2022-11 SA1#100 S1-223436 S1-223616 S1-223719 S1-223720 S1-223721 pCR on use case on supporting UAV pre-flight preparation New use case: Support UAV inflight monitoring New use case_VLoS_geofencing New use case on network-assisted UAV DAA New use case: 3GPP network as an information source to the UTM 0.2.0 2023-02 SA1#101 S1-230357 S1-230786 S1-230787 New use case on UAV flight route tracking at Rendezvous points pCR: Update on UTM pre-/in-flight operation support Update of the use case “Geofencing for Visual Line-of-Sight UAV missions” 0.3.0 2023-03 SA#99 SP-230221 MCC clean-up for presentation to SA 1.0.0 2023-05 SA1#102 S1-231258 S1-231608 S1-231610 S1-231611 S1-231614 S1-231617 S1-231622 S1-231638 S1-231762 S1-231773 pCR on Abbreviations section TR 22843 FS_UAV_ph3 quality improvement CR Use Case on UAV simultaneous traffic over two networks Use Case on UAV flying zones pCR on Use case Geofencing VLoS pCR on updating use case 5.2 Supporting UAV flight preparation pCR on use case on network-assisted UAV DAA Use Case on UAV traffic over two alternative networks pCR on FS-UAV-Ph3_Overview section pCR on consolidated requirements for UAV Phase3 1.1.0 2023-06 SA#100 SP-230517 MCC clean-up for approval by SA 2.0.0 2023-06 SA#100 SP-230517 Raised to v.19.0.0 by MCC following approval by SA 19.0.0 2023-09 SA#101 SP-231027 0001 C Removal of Editor's Notes in TR 22.843 on 5GS to UTM exposure of location 19.1.0 2023-09 SA#101 SP-231027 0003 F Update use case 5.3 Geofencing for Visual Line-of-Sight UAV missions 19.1.0 2023-09 SA#101 SP-231027 0004 1 B pCR on updating use case 5.4 network-assisted UAV DAA 19.1.0 2023-09 SA#101 SP-231027 0002 1 D Quality improvements to TR22.843 19.1.0 2023-09 SA#101 SP-231027 0005 2 B Conclusions and Recommendations 19.1.0 2023-12 SA#102 SP-231410 0008 D Term update in use case 5.3 Geofencing for Visual Line-of-Sight UAV missions 19.2.0 2023-12 SA#102 SP-231410 0007 3 C CPR alignment with agreed CR for TS 22.125 19.2.0 |
6fd8bf2627f82827816dd131ee257054 | 22.848 | 1 Scope | The present document describes use cases and aspects related to interconnect of SNPNs as well as Scalable SNPN Interconnect with dynamic connections.
Potential service requirements are derived for these use cases and are consolidated in a dedicated chapter.
The report ends with recommendations regarding the continuation of the work.
NOTE: There is no requirement for a PLMN to enhance their interconnect with SNPNs or operate an identity provider. |
6fd8bf2627f82827816dd131ee257054 | 22.848 | 2 References | The following documents contain provisions which, through reference in this text, constitute provisions of the present document.
- References are either specific (identified by date of publication, edition number, version number, etc.) or non‑specific.
- For a specific reference, subsequent revisions do not apply.
- For a non-specific reference, the latest version applies. In the case of a reference to a 3GPP document (including a GSM document), a non-specific reference implicitly refers to the latest version of that document in the same Release as the present document.
[1] 3GPP TR 21.905: "Vocabulary for 3GPP Specifications". |
6fd8bf2627f82827816dd131ee257054 | 22.848 | 3 Definitions of terms, symbols and abbreviations | |
6fd8bf2627f82827816dd131ee257054 | 22.848 | 3.1 Terms | For the purposes of the present document, the terms given in 3GPP TR 21.905 [1] and the following apply. A term defined in the present document takes precedence over the definition of the same term, if any, in 3GPP TR 21.905 [1].
SNPN Credential Provider: Entity within the 5G system that creates and manages identity information and provides authentication services for those identities for the purpose of accessing a SNPN
NOTE: The SNPN Credential Provider can also authorize access to a non-public network for a subscriber associated with an identity handled by this SNPN Credential Provider.
Interconnect: interaction between two SNPNs, or between a SNPN and a SNPN Credential Provider, allowing users subscribed to one SNPN (or to the SNPN Credential Provider) to use the other SNPN
Standalone Non-Public Network: an NPN that does not rely on network functions provided by a Public Land Mobile Network (PLMN) |
6fd8bf2627f82827816dd131ee257054 | 22.848 | 3.2 Abbreviations | For the purposes of the present document, the abbreviations given in 3GPP TR 21.905 [1] and the following apply. An abbreviation defined in the present document takes precedence over the definition of the same abbreviation, if any, in 3GPP TR 21.905 [1].
SNPN Standalone Non-Public Network |
6fd8bf2627f82827816dd131ee257054 | 22.848 | 4 Overview | The present document captures a set of use cases and potential service requirements related to the following topics:
1. Scalable SNPN Interconnect with dynamic connections (to support SNPN cellular hotspots).
2. Interconnect of SNPNs |
6fd8bf2627f82827816dd131ee257054 | 22.848 | 5 Use cases | |
6fd8bf2627f82827816dd131ee257054 | 22.848 | 5.1 Use case on Scalable SNPN Interconnect with dynamic connections | |
6fd8bf2627f82827816dd131ee257054 | 22.848 | 5.1.1 Description | This use case relates only to SNPNs that provide services in a similar way as provided by WLAN hotspots.
Today's PLMNs have a centralized database of IP addresses of all operator nodes that connect to the inter-PLMN IP backbone network, including AAA Servers/Proxies. This information is used for firewall and Border Gateway configuration. Signalling connections between VPLMN and HPLMN are long-lived (Diameter and HTTP-based N32f) and support bidirectional (inbound and outbound) signalling.
In contrast, there is no administrative entity that manages the interconnect of SNPNs on one hand and SNPN Credential Providers on the other hand. In addition, the number of interconnected SNPNs and SNPN Credential Providers is expected to be considerably greater than the number of PLMNs today. Given the large number of SNPN Credential Providers with which a SNPN needs to interconnect, it is not feasible to maintain permanently established signalling connections between them. Instead, the signalling connection needs to be established dynamically. Ideally, a signalling connection between SNPN A and SNPN Credential Provider B should be established only when a subscriber of SNPN Credential Provider B attempts to connect to SNPN A, and should be released when the last subscriber of SNPN Credential Provider B disconnects from SNPN A.
The dynamic establishment of a signalling connection between the SNPN and the SNPN Credential Provider raises several new issues, as follows:
- Outbound signalling issue: The SNPN might not know the IP address of the SNPN Credential Provider’s signalling endpoints and vice versa. In some cases, the SNPN Credential Provider may be operated using a cloud provider using dynamic IP address assignment.
- Inbound signalling issue: The SNPN and/or the SNPN Credential Provider may reside behind a firewall or a Network Address Translation (NAT) device.
- End-to-end signalling issue: The SNPN and/or the SNPN Credential Provider need to have assurance that they are indeed establishing a signalling connection with each other, and that the signalling connection is secure.
The issues related to dynamic establishment of a signalling connection between the SNPN and the SNPN Credential Provider are illustrated in Figure 5.1.1-1.
Figure 5.1.1-1: Issues with dynamic establishment of signalling connection between SNPN and SNPN Credential Provider |
6fd8bf2627f82827816dd131ee257054 | 22.848 | 5.1.2 Pre-conditions | 1) NetAbove is an Internet service provider that operates SNPNs and serves as a SNPN Credential Provider. As part of the subscription contract, NetAbove allows its subscribers to connect to SNPNs owned by a variety of companies (e.g., hospitality and convention centres, airports, government institutions, schools, restaurants, coffee shops, etc.).
2) NetAbove serves as a SNPN Credential Provider that can authenticate and authorize its subscriber when the subscriber attempts to connect to a SNPN.
3) NetAbove does not have a direct agreement with any of these SNPNs and instead relies on a third party that assists the interconnect of entities affiliated with this third party for the purpose of subscriber authentication and authorization.
4) Kaffa Koffee has launched a chain of coffee shops across the country. Each coffee shop is equipped with a small SNPN allowing customers to access Kaffa Koffee’s private entertainment system, in addition to providing Internet access.
5) Kaffa Koffee’s customers get free access to the entertainment system on the condition that they can assert an identity that can be authenticated by a supported SNPN Credential Provider.
6) A wide variety of companies can take the role of SNPN Credential Providers (e.g., mobile operators, other SNPNs). Kaffa Koffee does not have a direct agreement with any of these SNPN Credential Providers and instead relies on a third party that assists the interconnect of entities affiliated with this third party for the purpose of subscriber authentication and authorization.
7) BananaFed is a third party that assists the interconnect of entities affiliated with this third party for the purpose of subscriber authentication and authorization.
8) NetAbove and Kaffa Koffee both are affiliated with BananaFed. |
6fd8bf2627f82827816dd131ee257054 | 22.848 | 5.1.3 Service Flows | 1) Roy is a subscriber of NetAbove.
2) Roy is visiting a Kaffa Koffee shop and wants to access Kaffa Koffee’s famous entertainment system using his 5G-capable laptop.
3) Using the connection manager on his laptop, Roy selects his NetAbove subscription credentials for connection to Kaffa Koffee’s SNPN.
4) Kaffa Koffee’s SNPN has no direct signalling connection with NetAbove. In order to authenticate the request coming from Roy’s laptop, Kaffa Koffee’s SNPN turns to BananaFed, a third party, for assistance. Both NetAbove and Kaffa Koffee are affiliated with BananaFed.
5) NetAbove’s SNPN Credential Provider server resides in NetAbove’s administrative IP domain and is located behind a firewall.
6) With BananaFed’s assistance, the Kaffa Koffee’s SNPN is able to forward the outbound authentication request to a SNPN Credential Provider server that is owned by NetAbove, even though Kaffa Koffee’s SNPN has no prior knowledge of the IP address of any NetAbove’s SNPN Credential Provider servers.
7) With BananaFed’s assistance, the NetAbove’s SNPN Credential Provider server is able to receive the inbound authentication request originated by Kaffa Koffee’s SNPN even though it is located behind a firewall.
8) With BananaFed’s assistance, the Kaffa Koffee’s SNPN and NetAbove’s SNPN Credential Provider server get assurance that they are establishing a secure connection with the intended remote party.
9) Kaffa Koffee’s SNPN succeeds in authenticating Roy with the SNPN Credential Provider server at NetAbove.
10) NetAbove’s SNPN Credential Provider server desires to be informed when Roy disconnects from Kaffa Koffee’s SNPN. With BananaFed’s assistance, NetAbove’s SNPN Credential Provider server is able to forward the outbound subscription request to the SNPN operated by Kaffa Koffee, even though Kaffa Koffee’s SNPN is located behind a NAT device. |
6fd8bf2627f82827816dd131ee257054 | 22.848 | 5.1.4 Post-conditions | 1) Once Roy’s laptop is connected to the Kaffa Koffee’s SNPN network, Roy can enjoy access to Kaffa Koffee’s entertainment system and access to the Internet.
2) After some time, Roy disconnects gracefully from Kaffa Koffee’s SNPN and the disconnection event is signalled to NetAbove’s SNPN Credential Provider server. |
6fd8bf2627f82827816dd131ee257054 | 22.848 | 5.1.5 Existing features partly or fully covering the use case functionality | None. The existing interface for interconnect between SNPN and SNPN Credential Provider (aka Credentials Holder) relies on permanent signalling connection between the SNPN and the SNPN Credential Provider. |
6fd8bf2627f82827816dd131ee257054 | 22.848 | 5.1.6 Potential New Requirements needed to support the use case | [PR 5.1.6-001] Based on the Standalone Non-Public Network (SNPN) configuration, the 5G system shall support a mechanism for a SNPN to be able to interconnect with a large number of SNPN Credential Providers with which the SNPN might not have preconfigured information detailing the IP addresses used by these SNPN Credential Providers to interconnect with the SNPN.
[PR 5.1.6-002] Based on the Standalone Non-Public Network (SNPN) configuration, the 5G system shall support a mechanism for a SNPN Credential Provider to be able to interconnect with a large number of Standalone Non-Public Networks (SNPNs) with which the SNPN Credential Provider might not have preconfigured information detailing the IP addresses used by these SNPNs to interconnect with the SNPN Credential Provider.
[PR 5.1.6-003] Based on the Standalone Non-Public Network (SNPN) configuration, the 5G system shall support a mechanism for a SNPN to be able to determine how to connect to a SNPN Credential Provider capable of verifying the identity presented by a user attempting to connect to that SNPN.
[PR 5.1.6-004] Based on the Standalone Non-Public Network (SNPN) configuration, the 5G system shall support a mechanism for a SNPN to be able to securely interconnect with a SNPN Credential Provider in deployments where the required security information is not preconfigured.
[PR 5.1.6-005] Based on the Standalone Non-Public Network (SNPN) configuration, the 5G system shall support a mechanism for a SNPN to enable a SNPN Credential Provider to securely notify events (e.g., a user’s subscription ending) to the Standalone Non-Public Network (SNPN). |
6fd8bf2627f82827816dd131ee257054 | 22.848 | 5.2 Use case on Interconnect between SNPNs | |
6fd8bf2627f82827816dd131ee257054 | 22.848 | 5.2.1 Description | SNPNs are expected to be deployed in government offices, banks, hotels, hospitals, schools, etc. in an administrative area. Each SNPN exposes its localized services.
Interconnect between different SNPN, subject to appropriate agreements, can make access to localized services seamless. Namely, when SNPNs, deployed by different administrations, are interconnected with each other and their services exposed, local services of one SNPN can be accessed via the other SNPNs. For example, a medical service, instantiated on a SNPN deployed in a central hospital, can be exposed to distributed family clinics. Also, an identity created in a family clinic can be utilized at the central hospital.
A benefit of the interconnect between SNPN can also be seen when a disaster happens, where PLMN-based service become unavailable due to congestion caused by the disaster in the area. On the other hand, as SNPNs are basically isolated from the PLMNs, the SNPNs are not impacted by the congestion in the PLMNs. This is, for example, good for city governments because the city government can keep their public services over the interconnected SNPNs.
In the context of interconnected SNPNs, identity providers authenticate individual users in SNPNs and authorize them to consume services on the SNPNs.
Figure 5.2.1-1: Interconnect between SNPNs
Assume that multiple SNPNs, for example, SNPN-1, SNPN-2 and SNPN-3, are deployed at different places. In Figure 5.2.1-2, SNPN-1 is the central SNPN and SNPN-2 and SNPN-3 are branches. It is also expected that there is a backbone/transport network to provide interconnectivity among SNPN-1, SNPN-2 and SNPN-3. The backbone network could be a public network (e.g. The internet) or a private/ managed network (e.g. enterprise network owning the SNPNs).
Figure 5.2.1-2: Overview of interconnect
Following interconnectivity can be considered but the list is not exhaustive.
1. Point-to-point interconnectivity that connects a point with another point.
2. Hub & Spoke interconnectivity that connects a root point with other leaf points.
3. Gateway interconnectivity that connects multiple points with other multiple points with a full mesh topology
In a case where a pair of point-to-point interconnectivity in Figure 5.2.1-3 is deployed between SNPN-1 and SNPN-2 and between SNPN-2 and SNPN-3, the interconnectivity has 6 interconnectivity endpoints. Each of SNPN-1, SNPN-2 and SNPN-3 uses 2 interconnectivity endpoints to reach to other SNPNs. As for a traffic destined from a SNPN to another SNPN, each SNPN should manage which route is reached to another a SNPN (e.g. EP12 hosting SNPN-1 supports a point-to-point connectivity to EP21 hosting SNPN-2).
Figure 5.2.1-3: point-to-point connectivity
In a case where hub & spoke interconnectivity in Figure 5.2.1-4 is deployed among SNPN-1, SNPN-2 and SNPN-3, internal traffic among interconnectivity endpoints transferred via the root endpoint. For example, a central hospital provides SNPN-1, whereas family clinics #2 and #3 are consumers of SNPN-2 and SNPN-3, respectively. As assumed that SNPN-1 is the central SNPN, SNPN-1 can also provide identity services. For example, the internal traffic from EP-3L hosting SNPN-3 is transferred to SNPN-1 via EP-1R and then SNPN-1 routes the traffic from EP-1R to EP-2L hosting SNPN-2. It should be noted that an identity provider service could also be located in the external data network connecting to SNPN-1.
Figure 5.2.1-4: Hub & spoke connectivity
In a case where gateway interconnectivity in Figure 5.2.1-5 is deployed among SNPN-1, SNPN-2 and SNPN-3, internal traffic among interconnectivity endpoints transferred between each interconnectivity endpoint. Each SNPN can communicate with each other via the gateway interconnect service. For example, services such as an identity provider service could directly be interfaced with the interconnectivity. In this case, the identity provider service can seamlessly be consumed via any SNPNs and an identity assigned with a user can be authenticated and authorized in any SNPN.
Figure 5.2.1-5: Gateway connectivity
An interconnectivity (e.g. point-to-point, hub & spoke, gateway, etc.) is, for example, designed by a producer of the interconnect. The interconnect describes a collection of interconnectivity endpoints and how to access to the interconnectivity endpoints from SNPN-1, SNPN-2 and SNPN-3.
In general, the interconnect information is provided by a producer to consumers. For example, there is a case where the backbone network operator is a producer of the interconnect and SNPN-1, SNPN-2 and SNPN-3 are consumers. In another case, SNPN-1 is a producer of the interconnect and SNPN-2 and SNPN-3 are consumers. The interconnect information may be exchanged between the producer and consumers. |
6fd8bf2627f82827816dd131ee257054 | 22.848 | 5.2.2 Pre-conditions | It is assumed that a hub & spoke model is deployed in a central hospital and family clinics for this service flow. SNPN-1 is the root, deployed in central hospital and SNPN-2 and SNPN-3 are leaves, deployed in family clinics #2 and #3, respectively. SNPN-1, SNPN-2 and SNPN-3 are connected to the backbone network and they are at least IP reachable with each other. As such, this use case discusses a case of managed interconnect between different SNPNs. |
6fd8bf2627f82827816dd131ee257054 | 22.848 | 5.2.3 Service Flows | 1. An interconnect for a hub & spoke model is designed by a producer of SNPN-1, deployed in the central hospital. The interconnect describes a collection of interconnectivity endpoints for the root and branches and how to access to the interconnectivity endpoints from SNPN-1, SNPN-2 and SNPN-3.
2. The interconnect provided by SNPN-1 to SNPNs deployed in the family clinics, i.e. SNPN-2 and SNPN-3.
3. The management system of SNPN-1 selects an endpoint within 5GS of SNPN-1 to an interconnectivity endpoint (root) and then prepares an access link to the interconnectivity endpoint.
4. The management system of SNPN-2 selects an endpoint within 5GS of SNPN-2 to an interconnectivity endpoint (leaf) and then prepares an access link to the interconnectivity endpoint.
5. The management system of SNPN-3 selects an endpoint within 5GS of SNPN-3 to an interconnectivity endpoint (leaf) and then prepares an access link to the interconnectivity endpoint. |
6fd8bf2627f82827816dd131ee257054 | 22.848 | 5.2.4 Post-conditions | The interconnect between SNPN-1, SNPN-2 and SNPN-3 is realised using the selected topology model and 5GS endpoints, and then the SNPNs can exchange data. |
6fd8bf2627f82827816dd131ee257054 | 22.848 | 5.2.5 Existing features partly or fully covering the use case functionality | Further existing features that partly or fully cover the use case functionality are FFS. |
6fd8bf2627f82827816dd131ee257054 | 22.848 | 5.2.6 Potential New Requirements needed to support the use case | [PR 5.2.6-001] Subject to SNPNs operators’ agreement, the 5G system shall be able to support mechanisms to enable interconnect between standalone non-public networks.
[PR 5.2.6-002] Subject to SNPNs operators’ agreement, the 5G system shall be able to support mechanisms for a selection of 5GS endpoints for SNPNs’ interconnect. |
6fd8bf2627f82827816dd131ee257054 | 22.848 | 5.3 Use case on interconnection of standalone naval non-public networks | |
6fd8bf2627f82827816dd131ee257054 | 22.848 | 5.3.1 Description | This use case deals with self-organizing, broadband, low latency, connectivity solutions between ships in a fleet. It can apply to different sectors including military, coast guard fleet, or fleet of fishing vessels… in short to any situation where intra-fleet communication and coordination is needed.
In this use case, 5G systems are deployed on each ship within a fleet, where each system is a fully autonomous communication network enabling inter-ship and intra-ship communication. All vessels in the fleet eventually form a group of 5G standalone non-public networks connected altogether in an inherently variable topology. An example of such an interconnected group of standalone non-public networks is depicted on the Figure 5.3.1-1 below:
Figure 5.3.1-1: Example of interconnected group of standalone non-public networks
All standalone non-public networks are interconnected together through wireless links, preferably 5G links operating in the same frequency band as the 5G radio links connecting the terminals and the SNPNs (access network). In addition, as the group of interconnected SNPNs can be deployed in operation theatres, or in difficult sea conditions, it is important that its capacity be preserved even in the event of the loss of one or more vessels.
Each SNPN in the fleet can both be connected to the other ships through wireless links and offers connectivity in the immediate proximity of the ship where it is deployed. A user on a ship can consume any of the services hosted by the other vessels composing the fleet.
This system will finally have the following characteristics:
- Average size of targeted fleets: 5 or 6 ships
- Self-organized group of SNPNs with interconnection links established dynamically and according to the position of the ships in relation to each other’s and the quality of the radio links
- Dynamic routing according to the quality of access and interconnection links
- Full-duplex communication between ships
- Mobility of terminals from one ship (one SNPN) to another (another SNPN) without loss of communication
- Differentiated quality of service management
- Very high resilience, the connectivity between the set of SNPNs must remain operational during ship engagements, regardless of sea conditions and number of remaining vessels.
- Performances:
- 40 km range in line of sight for the interconnection links
- 1-10 Mbit/s in 5MHz bandwidth
- Latency << 100ms |
6fd8bf2627f82827816dd131ee257054 | 22.848 | 5.3.2 Pre-conditions | User A terminal is loaded with a USIM that is registered to SNPN A.
SNPN A has local coverage on-board ship A and SNPN B has coverage on-board ship B.
Service B is offered by an application B instantiated on ship B. Application B has 5G connectivity through SNPN B.
User A has needed credentials to access service B. |
6fd8bf2627f82827816dd131ee257054 | 22.848 | 5.3.3 Service Flows | 1) User A on-board ship A and attached to the on-board SNPN A establishes a PDU session to access service B.
2) In the process of establishing the PDU session, the 5G system identifies that communication with service B shall be routed from SNPN A to SNPN B through SNPN D to fulfil the quality of service required by service B. The 5G system configures all resources on this path according to service B quality of service requirements.
3) Later on, ships have moved relatively from each other’s, based on metrics collected on the wireless interconnection links, the 5G system now identifies that the communication with service B shall be routed through SNPN C. The call is handed over to the new path without the user A even noticing it.
4) Later on, unfortunately, SNPN C is no more available (or goes out of reach) and is replaced by a helicopter embedding an ad hoc SNPN. As part of a discovery process, the new SNPN in the helicopter identifies and authenticates itself to those SNPNs in range. The new route now comprising the SNPN on board the helicopter is evaluated against quality of service requirements and eventually selected by the 5G system for user A PDU session.
5) Meanwhile, while the helicopter is being launched, the 5G system realizes that the communication path using SNPN C is no more available and, as a result, User A communication to service B is reconfigured to the initial communication path through SNPN D. The communication is experiencing quality issues until the communication path is switched over to the one involving the SNPN on board the helicopter.
6) The communication with service B is terminated by user A and all resources used for the communication between SNPN A and SNPN B are released. |
6fd8bf2627f82827816dd131ee257054 | 22.848 | 5.3.4 Post-conditions | User A is attached to SNPN A. |
6fd8bf2627f82827816dd131ee257054 | 22.848 | 5.3.5 Existing features partly or fully covering the use case functionality | None |
6fd8bf2627f82827816dd131ee257054 | 22.848 | 5.3.6 Potential New Requirements needed to support the use case | [PR 5.3.6-001] Based on SNPN configuration and subject to SNPN operator’s policy, the 5G system shall be able to support interconnect of standalone non-public networks through links established on a temporary basis forming a group of interconnected Standalone Non-Public Networks with a variable topology.
[PR 5.3.6-002] Based on SNPN configuration and subject to SNPN operator’s policy, the 5G system shall be able to support routing of user data through interconnected Standalone Non-Public Networks.
[PR 5.3.6-003] Based on SNPN configuration and subject to SNPN operator’s policy, the 5G system shall be able to support discovery, addition and authentication, or detachment of Standalone Non-Public Networks in a group of interconnected Standalone Non-Public Networks.
[PR 5.3.6-004] Based on SNPN configuration and subject to SNPN operator’s policy, the 5G system shall be able to support per UE communication, the establishment, reconfiguration and release of user data transmission path between two Standalone Non-Public Networks in a group of interconnected Standalone Non-Public Networks.
[PR 5.3.6-005] Based on SNPN configuration and subject to SNPN operator’s policy, the 5G system shall be able to support a user subscribed to one Standalone Non-Public Network to consume a service offered by another Standalone Non-Public Network in a group of interconnected Standalone Non-Public Networks.
[PR 5.3.6-006] Based on SNPN configuration and subject to SNPN operator’s policy, the 5G system shall be able to support a mechanism to exploit interconnection links metrics (e.g. jitter, latency, packet loss, capacity) as a criterion to select a suitable route between two distant Standalone Non-Public Networks in a group of interconnected Standalone Non-Public Networks. |
6fd8bf2627f82827816dd131ee257054 | 22.848 | 5.4 Use case on mobility in a group of interconnected standalone naval non-public networks | |
6fd8bf2627f82827816dd131ee257054 | 22.848 | 5.4.1 Description | This use case deals with self-organizing, broadband, low latency, connectivity solution between ships in a fleet. It is an extension of the use case 5.3 on interconnection of standalone naval non-public networks.
This network architecture, unusual in the world of public cellular networks, will however have to support all the mobility functions supported by terrestrial networks. In particular, a terminal will have to be able to move freely from one vessel to another, either in standby or while being actively involved in a communication. This movement should not be perceptible to the user.
The situation described in this use case is depicted on the Figure 5.4.1-1 below:
Figure 5.4.1-1: Example of mobility in a group of interconnected standalone naval non-public networks |
6fd8bf2627f82827816dd131ee257054 | 22.848 | 5.4.2 Pre-conditions | User A terminal is loaded with a USIM that is registered to SNPN A.
SNPA A has local coverage on-board ship A and SNPN B has coverage on-board ship B.
Service B is offered by an application B hosted on ship B. Application B has 5G connectivity through SNPN B.
User A has needed credentials to access service B. |
6fd8bf2627f82827816dd131ee257054 | 22.848 | 5.4.3 Service Flows | 1) User A on-board ship A and attached to the on-board SNPN A establishes a PDU session to access service B.
2) User A takes a speed boat, equipped with necessary communication equipment to deploy locally SNPN around the speed boat and to enable connectivity with any vessels in the fleet through wireless links. Once it is deployed, the on-board SNPN selects the SNPNs in range, identifies and authenticates itself to those SNPNs. The SNPN on board the speed boat is then added in the group of interconnected SNPNs through wireless links.
3) As the speed boat moves away from ship A, radio measurements made by user A terminal trigger the 5G system to handover the ongoing communication from SNPN A to the SNPN on-board the speed boat.
4) At this point, user A terminal, although under radio coverage of the SNPN in the speed boat, keeps consuming service B through a communication path made of speed boat SNPN + SNPN A + SNPN D + SNPN B.
5) Somewhere along the path from ship A to ship B, as radio conditions on the 5G link to ship A worsen and as the speedboat approaches ship D, the SNPN on-board the speed boat selects SNPN D as a new anchoring point to the group of interconnected standalone non-public networks. Based on the new resulting topology, the 5G system then seamlessly reconfigures the communication path from user A’s UE to service B so that it can still continue to fulfil QoS required by service B. The communication path is now made of speed boat SNPN + SNPN D + SNPN B.
6) Later on, as the speedboat approaches ship B, the on-board SNPN directly connects to SNPN B in the same anchoring point selection and communication path switching process as previously described to reduce communication latency.
7) From the point in time user A gets on board the speedboat and gets to ship B, his communication with service B is maintained. There is no noticeable alteration of the communication along the path from ship A to ship B.
8) When user A arrives at his destination on ship B, his communication is seamlessly handed over from the SNPN on-board the speedboat to one of the cells on SNPN B. Once connected to SNPN B, he can still access to service B and other local services according to authorizations he’s entrusted with. |
6fd8bf2627f82827816dd131ee257054 | 22.848 | 5.4.4 Post-conditions | User A is attached to SNPN B and still in communication with service B. |
6fd8bf2627f82827816dd131ee257054 | 22.848 | 5.4.5 Existing features partly or fully covering the use case functionality | None |
6fd8bf2627f82827816dd131ee257054 | 22.848 | 5.4.6 Potential New Requirements needed to support the use case | [PR 5.4.6-001] Based on SNPN configuration and subject to SNPN operator’s policy, the 5G system shall be able to support discovery, addition and authentication, or detachment of Standalone Non-Public Networks in a group of interconnected Standalone Non-Public Networks without interrupting ongoing communications.
[PR 5.4.6-002] Based on SNPN configuration and subject to SNPN operator’s policy, the 5G system shall be able to support service continuity of UE communications between interconnected Standalone Non-Public Networks with overlapping radio coverage.
[PR 5.4.6-003] Based on SNPN configuration and subject to SNPN operator’s policy, the 5G system shall enable packet loss to be minimized during mobility of a UE between interconnected Standalone Non-Public Networks with overlapping radio coverage for some or all connections associated with this UE.
[PR 5.4.6-004] Based on SNPN configuration and subject to SNPN operator’s policy, the 5G system shall minimize communication service interruption time during mobility of a UE between interconnected Standalone Non-Public Networks with overlapping radio coverage for some or all connections associated with this UE. |
6fd8bf2627f82827816dd131ee257054 | 22.848 | 5.5 Use case on quality of service differentiation and resource prioritization in a group of interconnected standalone naval non-public networks | |
6fd8bf2627f82827816dd131ee257054 | 22.848 | 5.5.1 Description | This use case deals with self-organizing, broadband, low latency, connectivity solution between ships in a fleet. It is an extension of the use case 5.3 on interconnection of standalone naval non-public networks.
In this use case, several users located on different ships are involved in a real time mission critical communication while the interconnection links between the Standalone Non-Public Networks inside the group of interconnected Standalone Non-Public Networks experiences heavy load.
The situation described in this use case is depicted on the Figure 5.5.1-1 below:
Figure 5.5.1-1: Example of critical communication in a group of interconnected standalone naval non-public networks |
6fd8bf2627f82827816dd131ee257054 | 22.848 | 5.5.2 Pre-conditions | All User Equipment are loaded with a USIM and are registered to their local SNPN.
A MCX service is available onboard ship A (called MCX A in the following section)
All users involved in the use case have needed credentials to access MCX A service. |
6fd8bf2627f82827816dd131ee257054 | 22.848 | 5.5.3 Service Flows | 1) User A requests a real time video streaming involving a MCX server located onboard ship A to users B and D respectively located on ship B and D.
2) If the requested quality of service can be reached, users A, B, D on-board ships A, B, D, and attached to the on-board SNPNs A, B, D, can establish a PDU session to access the video streaming service provided by the MCX server on ship A.
3) While establishing the user B’s PDU session, the 5G system identifies that communication shall be routed from SNPN A to SNPN B through SNPN C to fulfil the quality of service required by the MCX service A. The 5G system configures all resources on this path according to MCX service A’s quality of service requirements.
4) While establishing the user D’s PDU session, the 5G system identifies that communication can be directly routed from SNPN A to SNPN D. The 5G system configures all resources on this path according to MCX service A’s quality of service requirements.
5) At this point in time, users A, B, D consume the real time video service hosted by the MCX server A and the 5G system constantly monitors the achieved end to end quality of service with regards to the application known requirements.
6) The delivered end to end quality of service changes (due to e.g. path loss increase on the interconnection links), thus triggering a notification from the 5G system to the MCX application which decides to lower the video resolution to match the end to end connection capacity and maintain an acceptable experience for the users.
7) Meanwhile, lower priority communications are setup inside the fleet and start to heavily load all the interconnection links.
8) The 5G system ensure that the critical video streaming communication between users A, B, D has higher priority related to the other ongoing data exchanges which results in user A, B, D continuing experiencing a good quality MCX service A. |
6fd8bf2627f82827816dd131ee257054 | 22.848 | 5.5.4 Post-conditions | Users A, B, D are attached to SNPN A, B, D and still in communication with MCX service A. |
6fd8bf2627f82827816dd131ee257054 | 22.848 | 5.5.5 Existing features partly or fully covering the use case functionality | None |
6fd8bf2627f82827816dd131ee257054 | 22.848 | 5.5.6 Potential New Requirements needed to support the use case | [PR 5.5.6-001] Based on SNPN configuration and subject to SNPN operator’s policy, the 5G system shall be able to support prioritization of resources for a service offered by a Standalone Non-Public Network that is consumed by users attached to other interconnected Standalone Non-Public Networks.
[PR 5.5.6-002] Based on SNPN configuration and subject to SNPN operator’s policy, the 5G system shall enable a SNPN to monitor the currently delivered end to end QoS for a service hosted by this SNPN and consumed by one or several users attached to other Standalone Non-Public Networks inside a group of interconnected Standalone Non-Public Networks.
[PR 5.5.6-003] Based on SNPN configuration and subject to SNPN operator’s policy, the 5G system shall be able to provide event notification upon detecting that the requested end to end QoS level cannot be met for a service hosted by this SNPN and consumed by one or several users attached to other Standalone Non-Public Networks inside a group of interconnected Standalone Non-Public Networks. |
6fd8bf2627f82827816dd131ee257054 | 22.848 | 6 Consolidated requirements | |
6fd8bf2627f82827816dd131ee257054 | 22.848 | 6.1 Groups of interconnected SNPNs | The potential requirements for enabling groups of interconnected SNPNs correspond to the use cases in clauses 5.2 to 5.5 and are listed in Table 6.1-1.
A group of interconnected SNPNs consists of one or more pairs of interconnected SNPNs. The determination of whether a given SNPN can or should be able to participate within a given group of interconnected SNPNs is based on prior arrangement and out of band configuration. How these arrangements and configuration are made is outside the scope of 3GPP and the 5G network.
Table 6.1-1 – Consolidated Requirements for groups of interconnected SNPNs
CPR #
Consolidated Potential Requirement
Original PR #
Comment
CPR 6.1-1
Based on SNPN configuration and subject to SNPN operator's policy, the 5G network shall be able to support mechanisms to enable interconnect between two or more SNPNs
NOTE: The interconnect can be established on temporary basis forming a group of interconnected SNPNs with variable topologies.
[PR 5.2.6-001]
[PR 5.2.6-002]
[PR 5.3.6-002]
[PR 5.3.6-001]
[PR 5.3.6-003]
CPR 6.1-2
Based on SNPN configuration and subject to SNPN operator’s policy, the 5G network shall be able to support prioritization of resources for a service offered by a SNPN that is consumed by users attached to other interconnected SNPNs
[PR 5.5.6-001]
CPR 6.1-3
Based on SNPN configuration and subject to SNPN operator’s policy, the 5G network shall be able to support service continuity for UEs that are moving between interconnected SNPNs
NOTE: It is assumed the interconnected SNPNs have overlapping radio coverage
[PR 5.4.6-002]
[PR 5.4.6-003]
[PR 5.4.6-004] |
6fd8bf2627f82827816dd131ee257054 | 22.848 | 6.2 Consolidated requirements for enabling of SNPN cellular hotspots | The potential requirements for enabling of SNPN cellular hotspots correspond to the use case in clause 5.1 (Scalable SNPN Interconnect with dynamic connections) and are listed in Table 6.2-1.
The requirements for SNPN cellular hotspots are intended to enable support of connectivity hotspots based on 3GPP 5G network that provide services in a similar way as provided by WLAN hotspots. Charging requirements are considered out of scope for this functionality.
Table 6.2-1 – Consolidated requirements for enabling of SNPN cellular hotspots
CPR #
Consolidated Potential Requirement
Original PR #
Comment
CPR 6.2-1
Based on the Standalone Non-Public Network (SNPN) configuration, the 5G network shall support a mechanism for an SNPN to be able to interconnect with a large number of SNPN Credential Providers with which the SNPN might not have preconfigured information detailing the IP addresses used by these SNPN Credential Providers to interconnect with the SNPN.
[PR 5.1.6-001]
Scalability
CPR 6.2-2
Based on the Standalone Non-Public Network (SNPN) configuration, the 5G network shall support a mechanism for an SNPN Credential Provider to be able to interconnect with a large number of Standalone Non-Public Networks (SNPNs) with which the SNPN Credential Provider might not have preconfigured information detailing the IP addresses used by these SNPNs to interconnect with the SNPN Credential Provider.
[PR 5.1.6-002]
Scalability
CPR 6.2-3
Based on the Standalone Non-Public Network (SNPN) configuration, the 5G network shall support a mechanism for an SNPN to be able to determine how to connect to an SNPN Credential Provider capable of verifying the identity presented by a user attempting to connect to that SNPN.
[PR 5.1.6-003]
CPR 6.2-4
Based on the Standalone Non-Public Network (SNPN) configuration, the 5G network shall support a mechanism for an SNPN to be able to securely interconnect with an SNPN Credential Provider in deployments where the required security information is not preconfigured.
[PR 5.1.6-004]
Security
CPR 6.2-5
Based on the Standalone Non-Public Network (SNPN) configuration, the 5G network shall support a mechanism for an SNPN to enable an SNPN Credential Provider to securely notify events (e.g., a user’s subscription ending) to the Standalone Non-Public Network (SNPN).
[PR 5.1.6-005]
Security |
6fd8bf2627f82827816dd131ee257054 | 22.848 | 7 Conclusions and recommendations | This technical report identifies several use cases and potential new requirements related to the interconnect of SNPNs with two aspects: - Groups of interconnected SNPNs - Enabling of SNPN cellular hotspots (Scalable SNPN Interconnect with dynamic connections) The resulting service requirements have been consolidated in clause 6. It is recommended to consider the consolidated requirements identified in this TR as the baseline for the subsequent normative work. Annex A (informative): Change history Change history Date Meeting TDoc CR Rev Cat Subject/Comment New version 2023-05 SA1#102 S1-231156 TR skeleton 0.0.0 2023-05 SA1#102 S1-231343 Inclusion of agreed pCRs: S1-231157, S1-231500, S1-231775, S1-231637, S1-231628, S1-231636 0.1.0 2023-08 SA1#103 S1-232598 Inclusion of agreed pCRs: S1-232461, S1-232639, S1-232466, S1-232467, S1-232459 0.2.0 2023-11 SA1#104 S1-233263 Inclusion of agreed pCRs: S1-233480 0.3.0 2024-03 SA1#105 S1-240314 Inclusion of agreed pCRs: S1-240129, S1-240226, S1-240281 0.4.0 2024-03 SA#103 SP-240193 MCC clean-up for presentation for one-step approval 1.0.0 2024-03 SA#103 SP-240193 Raised to v.19.0.0 following one-step approval at SA#103 19.0.0 |
80eab521a98c34c5f064bf3f6541cd5c | 22.850 | 1 Scope | This study will investigate ongoing AI/ML work in TSG CT, TSG RAN and TSG SA Working Groups and identify instances of any potential misalignment and/or inconsistencies and provide information on any potential outcome to the respective WGs to resolve any issues with appropriate SA-level co-ordination as necessary.
The study is led by TSG SA in close collaboration and inputs from TSG CT and TSG RAN. |
80eab521a98c34c5f064bf3f6541cd5c | 22.850 | 2 References | The following documents contain provisions which, through reference in this text, constitute provisions of the present document.
- References are either specific (identified by date of publication, edition number, version number, etc.) or non‑specific.
- For a specific reference, subsequent revisions do not apply.
- For a non-specific reference, the latest version applies. In the case of a reference to a 3GPP document (including a GSM document), a non-specific reference implicitly refers to the latest version of that document in the same Release as the present document.
[1] 3GPP TR 21.905: "Vocabulary for 3GPP Specifications".
[2] 3GPP TR 21.918: "Summary of Rel-18 Work Items".
[3] 3GPP TR 38.843: "Study on Artificial Intelligence (AI)/Machine Learning (ML) for NR air interface (Release 18)".
[4] 3GPP TR 23.700-84: "Study on Core Network Enhanced Support for Artificial Intelligence (AI) / Machine Learning (ML)".
[5] 3GPP TR 22.874: "5G System (5GS); Study on traffic characteristics and performance requirements for AI/ML model transfer".
[6] 3GPP TS 22.261: "Service requirements for the 5G system".
[7] 3GPP TR 23.700-82: "Study on application layer support for AI/ML services".
[8] 3GPP TS 23.288: "Architecture enhancements for 5G System (5GS) to support network data analytics services".
[9] 3GPP TS 28.105: "Management and orchestration; Artificial Intelligence/ Machine Learning (AI/ML) management".
[10] void
[11] 3GPP TS 38.300: "NR; NR and NG-RAN Overall description; Stage-2".
[12] 3GPP TR 26.927: "Study on Artificial Intelligence and Machine learning in 5G media services".
[13] 3GPP TS 38.401: "NG-RAN; Architecture description".
[14] 3GPP TS 38.420: "NG-RAN; Xn general aspects and principles".
[15] 3GPP TS 38.423: "NG-RAN; Xn Application Protocol (XnAP)".
[16] 3GPP TS 23.273: "5G System (5GS) Location Services (LCS); Stage 2".
[17] 3GPP TR 37.817: "Study on enhancement for data collection for NR and ENDC".
[18] 3GPP TR 28.908: "Study on Artificial Intelligence/Machine Learning (AI/ ML) management".
[19] 3GPP TR 28.858: "Study on Artificial Intelligence / Machine Learning (AI/ML) management; Phase 2".
[20] 3GPP TR 29.889: "Study on Protocol for AI Data Collection from UPF".
[21] 3GPP TR 22.876: "Study on AI/ML Model Transfer; Phase 2".
[22] 3GPP TS 23.501: "System architecture for the 5G System (5GS)".
[23] 3GPP TS 23.502: "Procedures for the 5G System (5GS)".
[24] 3GPP TS 23.503: "Policy and charging control framework for the 5G System (5GS); Stage 2".
[25] 3GPP TR 33.784: "Study on security aspects of core network enhanced support for Artificial Intelligence Machine Learning (AIML)".
[26] 3GPP TR 26.847: "Evaluation of Artificial Intelligence and Machine learning in 5G media services".
[27] 3GPP TS 28.552: "Management and orchestration; 5G performance measurements".
[28] 3GPP TS 32.425: "Telecommunication management; Performance Management (PM); Performance measurements Evolved Universal Terrestrial Radio Access Network (E-UTRAN)".
[29] 3GPP TS 28.554: "Management and orchestration; 5G end to end Key Performance Indicators (KPI)".
[30] 3GPP TS 28.541: "Management and orchestration; 5G Network Resource Model (NRM); Stage 2 and stage 3".
[31] 3GPP TS 32.254: "Telecommunication management; Charging management; Exposure function Northbound Application Program Interfaces (APIs) charging".
[32] 3GPP TS 32.291: "Telecommunication management; Charging management; 5G system, charging service; Stage 3".
[33] 3GPP TS 23.436: "Functional architecture and information flows for Application Data Analytics Enablement Service".
[34] 3GPP TS 23.482: "Functional architecture and information flows for AIML Enablement Service".
[35] 3GPP TS 23.434: "Service Enabler Architecture Layer for Verticals (SEAL); Functional architecture and information flows".
[36] 3GPP TS 23.558: "Architecture for enabling Edge Applications".
[37] 3GPP TS 29.122: "T8 reference point for Northbound APIs".
[38] 3GPP TS 29.513: "5G System; Policy and Charging Control signalling flows and QoS parameter mapping; Stage 3".
[39] 3GPP TS 29.514: "5G System; Policy Authorization Service; Stage 3".
[40] 3GPP TS 29.517: "5G System; Application Function Event Exposure Service; Stage 3".
[41] 3GPP TS 29.591: "5G System; Network Exposure Function Southbound Services; Stage 3".
[42] 3GPP TS 29.519: "5G System; Usage of the Unified Data Repository Service for Policy Data, Application Data and Structured Data for Exposure; Stage 3".
[43] 3GPP TS 29.520: "5G System; Network Data Analytics Services; Stage 3".
[44] 3GPP TS 29.521: "5G System; Binding Support Management Service; Stage 3".
[45] 3GPP TS 29.522: "5G System; Network Exposure Function Northbound APIs; Stage 3".
[46] 3GPP TS 29.523: "5G System; Policy Control Event Exposure Service; Stage 3".
[47] 3GPP TS 29.525: "5G System; UE Policy Control Service; Stage 3".
[48] 3GPP TS 29.508: "5G System; Session Management Event Exposure Service; Stage 3".
[49] 3GPP TS 29.551: "5G System; Packet Flow Description Management Service; Stage 3".
[50] 3GPP TS 29.552: "5G System; Network Data Analytics signalling flows; Stage 3".
[51] 3GPP TS 29.574: "5G System; Data Collection Coordination Services; Stage 3".
[52] 3GPP TS 29.575: "5G System; Analytics Data Repository Services; Stage 3".
[53] 3GPP TS 29.576: "5G System; Messaging Framework Adaptor Services; Stage 3".
[54] 3GPP TS 29.503: "5G System; Unified Data Management Services; Stage 3".
[55] 3GPP TS 29.504: "5G System; Unified Data Repository Services; Stage 3".
[56] 3GPP TS 29.505: "5G System; Usage of the Unified Data Repository services for Subscription Data; Stage 3".
[57] 3GPP TS 29.510: "5G System; Network function repository services; Stage 3".
[58] 3GPP TS 29.564: "5G System; User Plane Function Services; Stage 3".
[59] 3GPP TS 29.518: "5G System; Access and Mobility Management Services; Stage 3".
[60] 3GPP TS 29.536: "5G System; Network Slice Admission Control Services; Stage 3".
[61] 3GPP TS 29.571: "5G System; Common Data Types for Service Based Interfaces; Stage 3".
[62] 3GPP TS 38.214: "NR; Physical layer procedures for data".
[63] 3GPP TS 38.215: "NR; Physical layer measurements".
[64] 3GPP TS 38.331: "NR; Radio Resource Control (RRC); Protocol specification".
[65] 3GPP TS 38.305: "NG Radio Access Network (NG-RAN); Stage 2 functional specification of User Equipment (UE) positioning in NG-RAN".
[66] 3GPP TS 37.355: "LTE Positioning Protocol (LPP)".
[67] 3GPP TS 38.455: "NG-RAN; NR Positioning Protocol A (NRPPa)".
[68] 3GPP TS 38.133: "NR; Requirements for support of radio resource management".
[69] 3GPP TR 38.743: "Study on enhancements for Artificial Intelligence (AI)/Machine Learning (ML) for NG-RAN".
[70] 3GPP TR 38.744: "Study on Artificial Intelligence (AI)/Machine Learning (ML) for mobility in NR".
[71] 3GPP TS 28.104: “Management Data Analytics (MDA)”
[72] 3GPP TS 28.622: “Generic Network Resource Model (NRM) Integration Reference Point (IRP); Information Service (IS)” |
80eab521a98c34c5f064bf3f6541cd5c | 22.850 | 3 Definitions of terms and abbreviations | |
80eab521a98c34c5f064bf3f6541cd5c | 22.850 | 3.1 Terms | For the purposes of the present document, the terms given in TR 21.905 [1] and the following apply. A term defined in the present document takes precedence over the definition of the same term, if any, in TR 21.905 [1].
example: text used to clarify abstract rules by applying them literally. |
80eab521a98c34c5f064bf3f6541cd5c | 22.850 | 3.2 Abbreviations | For the purposes of the present document, the abbreviations given in TR 21.905 [1] and the following apply. An abbreviation defined in the present document takes precedence over the definition of the same abbreviation, if any, in TR 21.905 [1].
<ABBREVIATION> <Expansion> |
80eab521a98c34c5f064bf3f6541cd5c | 22.850 | 4 Background | In Rel-18 and Rel-19, most working groups in TSG SA, TSG CT and TSG RAN have already performed SIs and/or have WIs relating to the AI/ML topic. These activities address different usage scenarios and associated specific use cases exploiting AI/ML for the operation of the 3GPP System ranging from radio interface operations (e.g. beam management, positioning), NG-RAN operations (e.g. energy saving, load balancing), 5GC operations (e.g. assistance to QoS and policy control, prevention and mitigation of signalling storms, etc.), to network management & orchestration, media services, and application enablement aspects. In addition, some activities target the enhanced support of the 3GPP System for AI/ML-based applications and services which are themselves out of the scope of 3GPP.
With the complexity of the 3GPP systems and its operations and that of AI/ML, it is vital that the use of AI/ML in the operation of the 3GPP system (incl. related AI/ML model LCM) for any given use case be bound to specific principles, guidelines, design criteria, and requirements to safeguard the operation of the 3GPP System. This includes the capability to, e.g. fallback to non-AI/ML operation (i.e. not relying on inference process) whenever necessary not to negatively affect the NW and E2E performance.
This requires, as a minimum, the introduction of a common set of definitions to prevent any inconsistencies in the definition and use of AI/ML LCM across 3GPP WGs, to identify any misalignments/inconsistencies, and to communicate such inconsistencies to WGs for better alignment within 3GPP across different AI/ML related initiatives.
NOTE: AI/ML models and associated algorithms are certainly implementation specific and therefore out of scope of this study.
Editor's note: The study item does not impact ongoing studies and normative work for AI/ML across all SA/RAN/CT WGs for Rel-19. |
80eab521a98c34c5f064bf3f6541cd5c | 22.850 | 5 AI/ML related activities in all Working Groups | |
80eab521a98c34c5f064bf3f6541cd5c | 22.850 | 5.1 General | This clause will investigate and identify AI/ML related activities of all working groups of Rel-18 features and Rel-19 studies and work items in TSG CT, TSG RAN and TSG SA Working Groups, as given in the list in Table 5.1-1.
Table 5.1-1: List of Rel-18 and Rel-19 AI/ML activities in 3GPP working groups
Unique ID
Release
Title
Acronym
Approved SID/WID
TRs/TSs
SA WG1
920030
Rel-18
Stage 1 of AMMT
AIML_MT
SP-220440
TS 22.261 [6]
950008
Rel-19
Study on AI/ML Model Transfer Phase2
FS_AIML_MT_Ph2
SP-220439
TR 22.876 [21]
1000030
Rel-19
AI/ML Model Transfer Phase 2
AIML_MT_Ph2
SP-230514
TS 22.261 [6]
SA WG2
980019
Rel-18
System Support for AI/ML-based Services
AIMLsys
SP-231278
TS 23.501 [22]
TS 23.502 [23]
TS 23.503 [24]
TS 23.288 [8]
980020
Rel-18
Enablers for Network Automation for 5G - phase 3
eNA_Ph3
SP-230110
TS 23.501 [22]
TS 23.502 [23]
TS 23.503 [24]
TS 23.288 [8]
1020068
Rel-19
Study on Core Network Enhanced Support for Artificial Intelligence (AI)/Machine Learning (ML)
FS_AIML_CN
SP-231800
TR 23.700-84 [4]
1040033
Rel-19
Core Network Enhanced Support for Artificial Intelligence (AI)/Machine Learning (ML)
AIML_CN
SP-240991
TS 23.288 [8]
TS 23.501 [22]
TS 23.502 [23]
TS 23.503 [24]
TS 23.273 [16]
SA WG3
990042
Rel-18
Security aspects of enablers for Network Automation for 5G - phase 3
eNA_Ph3_SEC
Rel-18
SP-230155
TS 33.501 [22]
1030035
Rel-19
Study on security aspects of Core Network Enhanced Support for AIML
FS_AIML_CN_SEC
SP-240509
TR 33.784 [25]
SA WG4
950011
Rel-19
(started in Rel-18)
Feasibility Study on Artificial Intelligence (AI) and Machine Learning (ML) for Media
FS_AI4Media
SP-220328
TR 26.927 [12]
TR 26.847 [26]
SA WG5
990119
Rel-18
AI/ML management
AIML_MGT
SP-230335
TS 28.105 [9]
TS 28.552 [27]
TS 32.425 [28]
TS 28.554 [29]
TS 28.541 [30]
1020023
Rel-18
NEF Charging enhancement to support AI/ML in 5GS
AIMLsysNEF_CH
SP-231706
TS 32.254 [31]
TS 32.291 [32]
1020007
Rel-19
Study on AI/ML management - phase 2
FS_AIML_MGT_Ph2
SP-240965
TR 28.858 [19]
SA WG6
970036
Rel-18
Application Data Analytics Enablement Service
ADAES
SP-230275
TS 23.436 [33]
TS 23.434 [35]
TS 23.558 [36]
1010005
Rel-19
Study on application layer support for AI/ML services
FS_AIMLAPP
SP-231182
TR 23.700-82 [7]
1040075
Rel-19
Application enablement for AI/ML services
AIML_App
SP-241008
TS 23.482 [34]
TS 23.436 [33]
TS 23.434 [35]
TS 23.558 [36]
CT WG3
990008
Rel-18
CT WG3 aspects of System Support for AI/ML-based Services
AIMLsys
CP-230329
TS 29.122 [37]
TS 29.513 [38]
TS 29.514 [39]
TS 29.517 [40]
TS 29.591 [41]
TS 29.519 [42]
TS 29.520 [43]
TS 29.521 [44]
TS 29.522 [45]
990010
Rel-18
CT WG3 aspects of Enablers for Network Automation for 5G - phase 3
eNA_Ph3
CP-240079
TS 29.508 [48]
TS 29.122 [37]
TS 29.513 [38]
TS 29.517 [40]
TS 29.519 [42]
TS 29.520 [43]
TS 29.522 [45]
TS 29.523 [46]
TS 29.525 [47]
TS 29.551 [49]
TS 29.552 [50]
TS 29.574 [51]
TS 29.575 [52]
TS 29.576 [53]
TS 29.591 [41]
CT WG4
990008
Rel-18
CT WG4 aspects of AIML
Rel-18 CT WG4 aspects of AI/ML
CP-230329
TS 29.503 [54]
TS 29.504 [55]
TS 29.505 [56]
TS 29.510 [57]
TS 29.564 [58]
990010
Rel-18
CT WG4 aspects of Enablers for Network Automation for 5G - phase 3
eNA_Ph3
CP-240079
TS 29.503 [54]
TS 29.510 [57]
TS 29.518 [59]
TS 29.536 [60]
TS 29.564 [58]
TS 29.571 [61]
1040005
Rel-19
Study on Protocol for AI Data Collection from UPF
FS_PAIDC_UPF
CP-241025
TR 29.889 [20]
RAN WG1
1021093
Rel-19
Core part: Artificial Intelligence (AI)/Machine Learning (ML) for NR air interface
NR_AIML_airAir
RP-240774
TS 38.843 [3]
TS 38.300 [11]
TS 38.214 [62]
TS 38.215 [63]
TS 38.331 [64]
TS 38.305 [65]
TS 37.355 [66]
TS 38.455 [67]
TS 38.133 [68]
RAN WG2
1020084
Rel-19
Study on Artificial Intelligence (AI)/Machine Learning (ML) for mobility in NR
FS_NR_AIML_Mob
RP-240082
TR 38.744 [70]
RAN WG3
941010
Rel-18
Artificial Intelligence (AI)/Machine Learning (ML) for NG-RAN
NR_AIML_NGRAN-Core
RP-233441
TS 38.300 [11]
TS 38.401 [13]
TS 38.420 [14]
TS 38.423 [15]
1020083
Rel-19
Study on enhancements for Artificial Intelligence (AI)/Machine Learning (ML) for NG-RAN
FS_NR_AIML_NGRAN_enh
RP-240323
TR 38.743 [69]
RAN WG4
1022093
Rel-19
Perf. part: Artificial Intelligence (AI)/Machine Learning (ML) for NR air interface
NR_AIML_air-Perf
RP-240774
TS 38.133 [68]
Table 5.1-1 comprises 3GPP AI/ML related activities where the consumer of AI/ML models can be either the 3GPP System (including RAN, Core, OAM, and Application Enablement) itself to optimize 3GPP-specified functionalities and capabilities (i.e. equivalent to the common term “AI for Network”), or the applications and services that the 3GPP System provides support for (i.e. equivalent to the common term “Network for AI”).
Based on the above, the AI/ML related activities of 3GPP can be classified in the following two categories:
• ‘AI for Network’: Work to optimize the 3GPP System performance by employing AI/ML on 3GPP-specified functionalities and capabilities.
• ‘Network for AI’: Work to provide 3GPP System support for AI/ML-based applications and services.
Illustrative examples of 3GPP activities in Table 5.1-1 falling into the category ‘AI for Network’ include NR_AIML_air and AIML_MGT while ‘Network for AI’ activities include AIMLsys and AIML_App. In addition, certain items may address both ‘AI for Network’ and ‘Network for AI’ aspects, e.g. eNA_Ph3, AIML_CN.
Editor's note: Further review of Table 5.1-1 is needed to ensure completeness and identify relationships between different activities.
Editor's note: The classification of 3GPP activities in Table 5.1-1 to the above categories, and its benefit for the current study are FFS. |
80eab521a98c34c5f064bf3f6541cd5c | 22.850 | 5.2 AI/ML related activities in TSG SA & CT Working Groups | |
80eab521a98c34c5f064bf3f6541cd5c | 22.850 | 5.2.1 AI/ML related terminology | |
80eab521a98c34c5f064bf3f6541cd5c | 22.850 | 5.2.1.1 TSG SA WG2 | The following definitions are provided in clause 3 of TS 23.288 [8]:
- Analytics Accuracy Information: Represent a performance measure of an analytics ID provided by an NWDAF containing AnLF, which is composed of the number of correct predictions of the analytics ID out of all predictions and the corresponding number of samples.
- ML Model Accuracy Information: Represent a performance measure of a ML Model provided by an NWDAF containing MTLF, which is composed of the number of correct predictions by the ML Model out of all predictions and the corresponding number of samples.
- Analytics Feedback Information: Indicates that the consumer NF has taken action(s) influenced by the previously provided analytics, which may or may not affect the ground truth data.
The following definitions are provided in clause 5 of TS 23.288 [8]:
- Analytics logical function (AnLF): A logical function in NWDAF, which performs inference, derives analytics information (i.e. derives statistics and/or predictions based on Analytics Consumer request) and exposes analytics service i.e. Nnwdaf_AnalyticsSubscription or Nnwdaf_AnalyticsInfo.
- Model Training logical function (MTLF): A logical function in NWDAF, which trains Machine Learning (ML) models and exposes new training services (e.g. providing trained ML Model) as defined in clause 7.5 and clause 7.6.
The following definitions are provided in clause 3 of TR 23.700-84 [4]:
- Horizontal Federated Learning (HFL): a federated learning technique without exchanging/sharing local data set, wherein the local data set in different FL clients for local model training have the same feature space for different samples (e.g. UE IDs).
- Vertical Federated Learning (VFL): a federated learning technique without exchanging/sharing local data set, wherein the local data set in different VFL Participant for local model training have different feature spaces for the same samples (e.g. UE IDs).
- Label: A label is the training objective in supervised machine learning.
- VFL Server: An NWDAF or AF that integrates local training results, computes gradient information or loss information and send them to VFL client(s) for the local ML model update in VFL training process. It also coordinates the VFL training process by discovering and selecting VFL clients. In VFL inference process, The VFL server aggregates local inference results from VFL clients to generate the final VFL inference result and sends the final VFL inference result to the consumer. Only one VFL server may exist for each VFL process.
- VFL Client: An NWDAF or AF that holds the local dataset and performs local training and inference as asked by VFL Server. There can be multiple VFL Clients in VFL training and inference.
The following definitions are provided in clause 6.15 of TR 23.700-84 [4]:
- VFL active participant: A VFL function that owns part of an ML model for an analytic ID and knows the labels for the ML model. The active participant is the main function for training an ML model for an analytic ID.
- VFL passive participant: A VFL function that owns part of an ML model for an analytic ID but does not know the labels of the ML model but is able to collect local data for one or more features. |
80eab521a98c34c5f064bf3f6541cd5c | 22.850 | 5.2.1.2 TSG SA WG5 | The following definitions are provided in clause 3 of TS 28.105 [9]:
- ML model: a manageable representation of an ML model algorithm.
NOTE 1: An ML model algorithm is a mathematical algorithm through which running a set of input data can generate a set of inference output.
NOTE 2: An ML model algorithm is proprietary and not in scope for standardization and therefore not treated in this specification.
NOTE 3: An ML model may include metadata. Metadata may include e.g. information related to the trained model, and applicable runtime context.
- ML model training: a process performed by an ML training function to take training data, run it through an ML model algorithm, derive the associated loss and adjust the parameterization of that ML model iteratively based on the computed loss and generate the trained ML model.
- ML model initial training: a process of training an initial version of an ML model.
- ML model re-training: a process of training a previous version of an ML model and generate a new version.
NOTE 4: A new version of a trained ML model supports the same type of inference as the previous version of the ML model, i.e. the data type of inference input and data type of inference output remain unchanged between the two versions of the ML model, but parameter values might be different for the re-trained model.
- ML model joint training: a process of training a group of ML models.
- ML training function: a logical function with ML model training capabilities.
- AI/ML inference function: a logical function that employs an ML model to conduct inference.
- ML model testing: a process of evaluating the performance of an ML model using testing data different from data used for model training and validationtesting an ML model using testing data.
- ML model joint testing: a process of evaluating the performance of a group of ML models using testing data different from data used for model training and validation.
- ML testing function: a logical function with ML model testing capabilities.
- AI/ML inference: a process of running a set of input data through a trained ML model to produce set of output data, such as predictions.
NOTE 5: The inference represents the process to realize the AI capabilities by utilizing a trained ML model and other AI enablers if needed, hence the AI/ML prefix is used when referring to inference as compared to training and testing.
- AI/ML inference function: a logical function that employs trained ML model(s) to conduct inference.
- AI/ML inference emulation: running the inference process to evaluate the performance of an ML model in an emulation environment before deploying it into the target environment.
- ML model deployment: a process of making a trained ML model available for use in the target environment.
- ML model loading: a process of making a trained ML model available to an inference function.
- AI/ML activation: a process of enabling the inference capability of an AI/ML inference function.
- AI/ML deactivation: a process of disabling the inference capability of an AI/ML inference function.
The following definitions are provided in clause 3 of TR 28.858 [19]:
Distributed training: an ML training approach that distributes the training workload across multiple ML training functions.
Federated Learning: a distributed machine learning approach where the ML model is trained collaboratively by multiple ML training functions. This includes multiple FL clients, which perform training on local data, and one FL server, which aggregates model updates from the clients iteratively without exchanging data samples.
Fine-tuning: the process of training a pre-trained ML model with a changed or narrowed scope.
FL client: a training function that trains an ML model on local data and share only the model outcome with the FL server/FL Client, preserving data privacy.
FL Server: a function that aggregates the ML model outcomes from FL clients to produce a global ML model.
Horizontal Federated Learning (HFL): a federated learning technique without exchanging/sharing local data set, wherein the local data set in different FL clients for local model training have the same feature space for different samples.
ML Knowledge-based Transfer Learning: a technique where the knowledge gained from training of one or more ML models is applied or adapted to improve or develop another ML model.
Pre-training: the process of training an ML model on a dataset not specific to any type of inference.
Reinforcement Learning: a machine learning approach where an RL agent learns to make decisions by interacting with the environment and taking actions to maximize cumulative rewards (see NOTE 1).
NOTE 1: The examples of rewards could be improved resource allocation, reduced latency, or enhanced user experience.
Vertical Federated Learning (VFL): a federated learning technique without exchanging/sharing local data set, wherein the local data set in different VFL participant for local model training have different feature spaces for the same samples.
NOTE 1: The definitions of HFL and VFL referenced above are sourced from TR 23.700-84 [3]. This definition might be updated or modified in normative work.
NOTE 2: The definition of VFL is included for completeness and may need to be revisited with considerations of concrete use cases where VFL may be used in 5GS. |
80eab521a98c34c5f064bf3f6541cd5c | 22.850 | 5.2.1.3 TSG SA WG6 | The following definitions are provided in clause 3 of TR TS 23.482 [34]:
- ML model: According to TS 28.105 [9], mathematical algorithm that can be "trained" by data and human expert input as examples to replicate a decision an expert would make when provided that same information.
- ML model lifecycle: The lifecycle of an ML model aka ML model operational workflow consists of a sequence of ML operations for a given ML task / job (such job can be an analytics task or a VAL automation task). This definition is aligned with the 3GPP definition on ML model lifecycle according to 3GPP TS 28.105 [9].
- ML model training: According to TS 28.105 [9], ML model training includes capabilities of an ML training function or service to take data, run it through an ML model, derive the associated loss and adjust the parameterization of that ML model based on the computed loss.
- ML model inference: According to TS 28.105 [9], ML model training includes capabilities of an ML model inference function that employs an ML model and/or AI decision entity to conduct inference.
- AI/ML intermediate model: For federated learning, members need to train models for multiple rounds, intermediate models indicate the model which do not meet the required training rounds and/or meet the requirements of the federated training.
- AIMLE service: An AIMLE service is an AIMLE capability which aims assisting in performing or enabling one or more AIML operations.
- AI/ML client: an application layer entity (also referred as ML client) which is an AI/ML endpoint and performs client-side operations (e.g. related to the ML model lifecycle). Such AI/ML client can be a VAL client or AIML enabler client and may be configured e.g. to provide ML model training and inference locally e.g. at the VAL UE side.
- AIMLE client set identifier: an identifier of the set of selected AIMLE clients.
- AI/ML server: an application layer entity which is an AI/ML endpoint and performs server-side operations (e.g. related to the ML model lifecycle). Such AI/ML server can be a VAL server or AIML enabler server.
- FL member: An FL member or participant is an entity which has a role in the FL process. An FL member can be an FL client performing ML model training, or an FL server performing aggregation/collaboration for the FL process.
- FL client: An FL member which locally trains the ML model as requested by the FL server. Such FL client functionality can be at the network (e.g. AIMLE server with FL client capability) or at the device side (e.g. AIMLE client with FL client capability).
- FL server: An FL member which generates global ML model by aggregating local model information from FL clients.
- Split AI/ML operation pipeline: A Split AI/ML operation pipeline is a workflow for ML model inference in which AI/ML endpoints are organized and collaborate to process ML models in sequential stages, where processing at each stage involves ML model inference on the output of the previous stage. |
80eab521a98c34c5f064bf3f6541cd5c | 22.850 | 5.2.2 AI/ML related activities | Editor's note: Description clause will be further clarified whether to be based on WI summaries or WID objectives. |
80eab521a98c34c5f064bf3f6541cd5c | 22.850 | 5.2.2.1 Rel-18 SA WG1 WID - AI/ML model transfer in 5GS (AIML_MT) | |
80eab521a98c34c5f064bf3f6541cd5c | 22.850 | 5.2.2.1.1 Description | The objective of this work item is to specify performance requirements (for end-to-end latency, experienced data rate, communication service availability) and service requirements (for AI/ML QoS management, AI/ML model /data distribution/transfer, network performance and resource utilization monitoring/prediction) for 5GS to support the following AI/ML operations for various applications (e.g. image/speech recognition, media editing/enhancements, robot control, automotive):
- AI/ML operation splitting between AI/ML endpoints.
- AI/ML model/data distribution and sharing over 5G system.
- Distributed/Federated Learning over 5G system. |
80eab521a98c34c5f064bf3f6541cd5c | 22.850 | 5.2.2.1.2 Activities summary | Editor's note: This clause describes high-level AI/ML activities e.g. LCM for AI/ML, data collection/storage/exposure, model training/delivery/ (de)-activation/inference emulation, inference/storage/exposure, performance evaluation and accuracy monitoring. Clause(s) may be added to capture details.
The three application AI/ML operations in 5.2.2.1.1 the 5G system can support were specified in cl. 6.40.1 of TS 22.261 [6] as follows:
- AI/ML operation splitting between AI/ML endpoints: The AI/ML operation/model is split into multiple parts according to the current task and environment. The intention is to offload the computation-intensive, energy-intensive parts to network endpoints, whereas leave the privacy-sensitive and delay-sensitive parts at the end device. The device executes the operation/model up to a specific part/layer and then sends the intermediate data to the network endpoint. The network endpoint executes the remaining parts/layers and feeds the inference results back to the device.
- AI/ML model/data distribution and sharing: Multi-functional mobile terminals might need to switch the AI/ML model in response to task and environment variations. The condition of adaptive model selection is that the models to be selected are available for the mobile device. However, given the fact that the AI/ML models are becoming increasingly diverse, and with the limited storage resource in a UE, it can be determined to not pre-load all candidate AI/ML models on-board. Online model distribution (i.e. new model downloading) is needed, in which an AI/ML model can be distributed from a Network endpoint to the devices when they need it to adapt to the changed AI/ML tasks and environments. For this purpose, the model performance at the UE needs to be monitored constantly.
- Distributed/Federated Learning over 5G system: The cloud server trains a global model by aggregating local models partially-trained by each end devices. Within each training iteration, a UE performs the training based on the model downloaded from the AI server using the local training data. Then the UE reports the interim training results to the cloud server via 5G UL channels. The server aggregates the interim training results from the UEs and updates the global model. The updated global model is then distributed back to the UEs and the UEs can perform the training for the next iteration.
It is worth emphasizing that the above descriptions refer to AI/ML operations over the application layer. The service requirements and performance requirements for AI/ML model transfer over the application layer in 5GS with direct network connection are specified in cl. 6.40.2.1 and cl. 7.10.1 of TS 22.261 [6], respectively. |
80eab521a98c34c5f064bf3f6541cd5c | 22.850 | 5.2.2.2 Rel-19 SA WG1 SID - AI/ML Model Transfer Phase 2 (FS_AIML_MT_Ph2) | |
80eab521a98c34c5f064bf3f6541cd5c | 22.850 | 5.2.2.2.1 Description | The objective of this study is to explore new use cases and potential service and performance requirements to support efficient AI/ML operations using direct device connections. This includes:
- Distributed AI training and inference based on direct device connections, such as traffic KPIs, various QoS, and functional requirements for sidelink transmission.
- Considerations for charging and security aspects. |
80eab521a98c34c5f064bf3f6541cd5c | 22.850 | 5.2.2.2.2 Activities summary | Editor's note: This clause describes high-level AI/ML activities e.g. LCM for AI/ML, data collection/storage/exposure, model training/delivery/ (de)-activation/inference emulation, inference/storage/exposure, performance evaluation and accuracy monitoring. Clause(s) may be added to capture details.
In this study, TR 22.876 [21] described study the use cases with potential functional and performance requirements to support efficient AI/ML operations over the application layer using direct device connection for various applications e.g. auto-driving, robot remote control, video recognition, etc. The agreed activities which were progressed in normative phase are described in clause 5.2.2.3.2. |
80eab521a98c34c5f064bf3f6541cd5c | 22.850 | 5.2.2.3 Rel-19 SA WG1 WID - AI/ML Model Transfer Phase 2 (AIML_MT_Ph2) | |
80eab521a98c34c5f064bf3f6541cd5c | 22.850 | 5.2.2.3.1 Description | The objective of this work item is to specify KPI and functional requirements for 5GS to support the AIML data transfer by leveraging direct device connection under 5G network control. These objectives were derived based on outcome of Rel-19 study in SA WG1 that relates to how the 5GS supports the transmissions of AI/ML-based services over the application layer. The study addressed use cases and potential performance requirements for 5G system support of application layer Artificial Intelligence (AI)/Machine Learning (ML) model distribution and transfer (download, upload, updates, etc.), and identified traffic characteristics of AI/ML model distribution, transfer and training for various applications, e.g. video/speech recognition, robot control, automotive, other verticals. |
80eab521a98c34c5f064bf3f6541cd5c | 22.850 | 5.2.2.3.2 Activities summary | Editor's note: This clause describes high-level AI/ML activities e.g. LCM for AI/ML, data collection/storage/exposure, model training/delivery/ (de)-activation/inference emulation, inference/storage/exposure, performance evaluation and accuracy monitoring. Sub-clause(s) may be added to capture details.
The service requirements and performance requirements for AI/ML model transfer over the application layer in 5GS with direct device connection are specified in cl. 6.40.2.2 and cl. 7.10.2 of TS 22.261 [6], respectively. |
80eab521a98c34c5f064bf3f6541cd5c | 22.850 | 5.2.2.4 Rel-18 SA WG2 WID - Enablers for Network Automation for 5G - phase 3 (eNA_Ph3) | |
80eab521a98c34c5f064bf3f6541cd5c | 22.850 | 5.2.2.4.1 Description | The objective of this work item is to further enhance NWDAF, based on what has been specified in the previous releases to allow 5GS to support network automation. This work item focuses on architecture enhancement, new scenarios and the necessary inputs and outputs to the NWDAF based on the conclusions of the study in Rel-18. The work focuses on 10 key aspects as follows:
- Improve correctness of NWDAF analytics.
- NWDAF-assisted application detection.
- Data and analytics exchange in roaming case.
- Enhancements on Data collection and Storage.
- Enhancements on trained ML Model sharing.
- NWDAF-assisted URSP.
- Enhancements on QoS Sustainability analytics.
- Supporting Federated Learning in 5GC.
- Enhancement of NWDAF with finer granularity of location information.
- Interactions with MDAS/MDAF. |
80eab521a98c34c5f064bf3f6541cd5c | 22.850 | 5.2.2.4.2 Activities summary | Editor's note: This clause describes high-level AI/ML activities e.g. LCM for AI/ML, data collection/storage/exposure, model training/delivery/ (de)-activation/inference emulation, inference/storage/exposure, performance evaluation and accuracy monitoring. Clause(s) may be added to capture details.
5.2.2.4.2.1 AIML related LCM activities
Data collection/storage/exposure
Analysis of data collection activities as part of eNA, eNA_ph2 work.
- Data collection in 3GPP TS 23.288 [8] refers to data collected by the NWDAF and DCCF. The data are collected for the purpose of analytics generation and training of ML models. NWDAF/DCCF/MFAF collects data from NFs/AFs, OAM and UE application.
- Data Collected and Analytics output information may be stored at ADRF
As depicted in Figure 4.2.0-1 in TS 23.288, the 5G System architecture allows NWDAF to collect data from any 5GC NF. The NWDAF is allowed to collect data from any 5GC NF directly, and retrieve the management data from OAM by invoking OAM services as defined in clause 5.11 of TR 28.
As defined in clause 4.2.0 of TS 23.288, the NWDAF is allowed cdata from any 5GC NF using a DCCF or MFAF as defined in Figure 5A.3.2-1in TS 23.288, and retrieve the management data from OAM by invoking OAM services as defined in clause 5.11 of TR 28.871.As defined in clause 5A of TS 23.288, Each Event Notification received from a Data Source NF is sent to the DCCF which propagates it to all Data Consumers / Notification Endpoints specified by the Data Consumers or determined by the DCCF. As defined in clause 5B of TS 23.288, the ADRF offers services that enable a consumer to store, retrieve and delete data, analytics and ML Models. ML Model(s) may be stored in the ADRF by a consumer sending the ADRF a storage request containing the ML Model or ML Model address to be stored.
Analysis of Data Collection activities as part of eNA_ph3 work.
- Exposure of input data for analytics is allowed from VPLMN to HPLMN and vice versa.
- Data Processing enhancements at Data stored at ADRF.
AI/ML Model Training
Analysis AI/ML model training activities as part of eNA, eNA_ph2 work.
- MTLF trains an ML model for an Analytics output requested by an AnLF. Trained ML model is assigned an ML Model Identifier and may be stored in ADRF.
Analysis of AI/ML Model Training activities as part of eNA_ph3 work.
- Trained ML model sharing between different NWDAF vendors by defined ML Model Interoperability Indication/Information.
- Support of Model Training using Horizontal Federated Learning
AI/ML Model Inference
Analysis of AI/ML Model inference activities as part of eNA_ph3 work
- AnLF exposes analytics information to consumers (e.g. NFs). Analytics information may be stored at an ADRF.
- Analytics from multiple NWDAF can be aggregated
- Analytics content can be transferred between NWDAFs, including roaming scenarios.
Performance evaluation and accuracy monitoring
Analysis of performance evaluation and accuracy monitoring activities as part of eNA_Ph3 work:
- NWDAF (either AnLF or MTLF) supporting performance monitoring of a trained ML model.
5.2.2.4.2.2 AI/ML functional entities.
As part of eNA, eNA_ph2 and eNA_ph3 work the following functional entities have been defined.
- NWDAF (Network Data Analytics Function) is defined in TS 23.288 [8] with main function to generate analytics (statistics and/or predictions) for one or more network events. NWDAF is defined by two logical functions.
- AnLF (Analytics Logical Function): Derives and exposes analytics information (statistics or predictions)
- MTLF (Model Training Logical Function): Trains Machine Learning (ML) models and exposes new training services (e.g. providing trained ML Model)
- DCCF (Data Collection & Coordination Function) is defined in TS 23.288 [8] with main functionality for analytics collection from NWDAFs and data collection from multiple NF(s), AF and OAM.
- MFAF (Messaging Framework Adaptor NF) is defined in TS 23.288 [8] and is part of the DCCF architecture. MFAF offers 3GPP defined services that allow the 5GS to interact with a Messaging Framework
- ADRF (Analytics Data Repository Function) is defined in TS 23.288 [8] with main functionality for storing and retrieving collected data and analytics. |
80eab521a98c34c5f064bf3f6541cd5c | 22.850 | 5.2.2.5 Rel-18 SA WG2 WID - System Support for AI/ML-based Services (AIMLsys) |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.