Content for  TR 22.835  Word version:  18.2.0

As more network slices are deployed, it is likely that each UE can access to multiple network slices. While some network slice services can be simultaneously provided to the UE, there can be network slices which cannot simultaneously provide services to the UE. This is because there are multiple related factors for a network slice, e.g., enterprise use vs personal use, isolation requirement, general public use vs public safety use, frequency limitation, location restriction, etc.

Network slices are deployed by same geographical location but they are incompatible by configuration. Figure 5.3.2-1 shows the use case scenario where a RAN node connects to two network slices simultaneously. Slice M and Slice N have no shared/common core network nodes supporting these network slices due to network resources incompatibility in 5GC. For example, Slice M is used for public security and Slice N is used for Internet access. In the 5GC, dedicated network resources and network functionalities are separately customised for public security emergency service and video service to meet the isolation requirements, which ensures the independence of core network resources between different network slices. Accordingly, the two network slices are isolated and cannot be simultaneously provided to the UE.

In this Figure, it is assumed that

• Subscription and configuration:
  • UE A1 and A3 are authorized to access Slice M.
  • UE A2 and A3 are authorized to access Slice N.
  • For UE A3, it is configured which applications use which network slices.
• Deployment:
  • Slice N and Slice M are deployed at the same geographical location. Slice N and Slice M are isolated. Due to the network resources incompatibility in 5G core network, there are no shared/common core network nodes supporting the Slice N and Slice M.
  • NG-RAN is able to connect to both Slice M and Slice N.
  • Slice M and Slice N are provided by the same PLMN.

Following is the service flow for UE A1, A2 and A3 when the signaling from these UEs is initially routed toward nodes managing Slice N:

• UE A1, A2 and A3 select the NG-RAN and start registration.
• Network slices provide connectivity services to UE A1, A2 and A3.
• While the UE A3 is connected to Slice N, Slice M cannot provide the connectivity service to UE A3. The user traffic applicable to Slice N is transported.
• Later, user traffic configured to use slice M is generated in UE A3. While the same NG-RAN is used, the traffic is switched to Slice M.
• If the user prioritizes applications configured to use Slice N, as soon as applications configured to use Slice M finish, the UE A3 needs to move back to Slice N, to minimize the latency.

UE A1 is connected to Slice M. UE A2 is connected to Slice N. UE A3 switches between Slice N and Slice M, depending on e.g. ongoing applications.

There are related requirements on change of network slices based on services, as defined in clause 6.1.2.2 in TS 22.261: There are related requirements on impacts on other network slices caused by traffic and services from one network slice, as defined in clause 6.1.2.2 in TS 22.261: Based on clause 6.1 of TS 22.261 network slicing, the following service requirement is supported:

The following new requirements can be derived from this use case: [PR.5.3.6-1] [PR.5.3.6-2]